UNIVERSITY OF CINCINNATI
Date: October 7th, 2004
I, ______Yukari Tokuyama______, hereby submit this work as part of the requirements for the degree of: Doctor of Philosophy in: Cell and Molecular Biology It is entitled: NPM/B23: The Effector of CDK2 in the Control of Centrosome Duplication and mRNA Processing
This work and its defense approved by:
Chair: Kenji Fukasawa, Ph.D. Robert W. Brackenbury, Ph.D. Wallace S. Ip, Ph.D. Erik S. Knudsen, Ph.D. Yolanda Sanchez, Ph.D.
NPM/B23: THE EFFECTOR OF CDK2 IN THE CONTROL OF CENTROSOME DUPLICATION AND mRNA PROCESSING
A dissertation submitted to the
Division of Research and Advanced Studies of the University of Cincinnati
in partial fulfillment of the requirements for the degree of
DOCTORATE OF PHILOSOPHY (Ph.D.)
In the Department of Cell Biology, Neurobiology, and Anatomy of the College of Medicine
2004
by
Yukari Tokuyama
B.S. Texas A&M University, 1998
Committee Chair: Kenji Fukasawa, Ph.D. Robert W. Brackenbury, Ph.D. Wallace S. Ip, Ph.D. Erik S. Knudsen, Ph.D. Yolanda Sanchez, Ph.D.
Abstract
Nucleophosmin (NPM/B23) is a phosphoprotein predominantly localized in the
nucleolus, and is believed to function in ribosome assembly, cytonuclear shuttling, and
molecular chaperoning. Here we report novel functions of NPM/B23 in initiation of
centrosome duplication and mRNA processing. The centrosome, a major microtubule
organizing center of cells, directs the formation of bipolar mitotic spindles, which is
essential for proper chromosome segregation to daughter cells. Cancer cells frequently
contain abnormal numbers of centrosomes suggesting that centrosome amplification is
the major contributing factor for chromosome instability. It has been shown that cyclin-
dependent kinase 2 (CDK2) activity is required for initiation of centrosome duplication.
NPM/B23 is a substrate of CDK2/Cyclin E in initiation of centrosome duplication. Here
we identified that threonine 199 (Thr199) of NPM/B23 is the major phosphorylation site of CDK2/Cyclin E. NPM/B23 associates with single centrosomes and upon phosphorylation by CDK2/Cyclin E, NPM/B23 is lost from the centrosomes resulting in initiation of centrosome duplication. A non-phosphorylatable mutant of NPM/B23
(NPM/T199A) acts as a dominant negative, resulting in inhibition of centrosome duplication. Thus, NPM/B23 is a key regulator in initiation of centrosome duplication.
We further examined the role of CDK2/Cyclin E phosphorylation of NPM/B23 on
Thr199 (phospho-Thr199 NPM). We found that phospho-Thr199 NPM specifically
localizes to nuclear structures known as nuclear speckles, which consist of pre-mRNA
splicing factors. We found that phosphorylation on Thr199 of NPM/B23 enhances
NPM/B23’s RNA binding activity. Also, non-phosphorylated NPM/B23 binds with specific splicing factors, whereas phosphorylation of NPM/B23 decreases these
interactions. Moreover, we found that phosphorylation of NPM/B23 on Thr199 represses
pre-mRNA splicing. These findings suggest an interesting connection between cell cycle
progression and mRNA processing. We also found NPM/B23 to be ubiquitinated.
NPM/B23 is maximally ubiquitinated at the G1/S transition of the cell cycle, when
CDK2/Cyclin E is also maximally active. Furthermore, phosphorylation on Thr199 targets NPM/B23 for ubiquitination. Studies have shown that ubiquitin ligases such as
SCF and BRCA1-BARD1, localize to centrosomes, and ubiquitin-mediated proteolysis
plays an important role in centrosome duplication. Thus, it is possible that ubiquitin
modification of NPM/B23 is a key event in the centrosome duplication cycle.
Acknowledgments
This work would not have been possible without the help, support, and guidance of so many people. Especially, I am grateful to my advisor, Dr. Kenji Fukasawa, for his guidance and patience throughout these years. He has been a great advisor, strengthening and shaping me to become a scientist. I am thankful to my committee members for all their advise and keeping me on the right path.
The past and present members of the Fukasawa lab have made my graduate career enjoyable. I would especially like to thank Dr. Masaru Okuda who has taught me the basics when I first joined the lab. Also, I would like to thank Dr. Pheruza Tarapore and
Dr. Henning Horn for their countless help and support…thank you for putting up with me for all these years!
I would also like to thank all my friends who have supported and encouraged me throughout graduate school. I am most grateful to my friends Rick Flannery and Robin
Kuns for their friendship, support and encouragement. They have always been there for me in the toughest of times, without them I am not sure if I would have made it.
This dissertation is dedicated to my family (mom, dad and Mayumi). Thank you for your understanding, endless support and encouragement…words can’t really describe how grateful I am to have a wonderful family!
TABLE OF CONTENTS
Page List of Tables and Figures…………………………………………………………….2
Chapter I: Introduction………………………………………………………………..4 A. NPM/B23 as a multi-functional protein………………………………5 B. NPM/B23 functional domains………………………………………...7 C. NPM/B23 function is mediated by phosphorylation events…………..9 D. Centrosomes and the cell cycle……………………………………….10 E. Involvement of CDK2/Cyclin E in regulation of centrosome duplication……………………………………………..14 F. Nuclear Speckles……………………………………………………...15 G. Pre-mRNA splicing…………………………………………………...17 H. Ubiquitination……………………………………………...…………20 I. References………………………………………………………...…..22
Chapter II-IV: Results
Chapter II: Specific Phosphorylation of Nucleophosmin on Thr199 by Cyclin-dependent Kinase 2-Cyclin E and Its Role in Centrosome Duplication……………………………………………….....29 A. Abstract……………………………………………………………….30 B. Introduction…………………………………………………………...32 C. Materials and Methods………………………………………………..34 D. Results ………………………………………………………………...39 E. Discussion…………………………………………………………….47 F. References…………………………………………………………….52 G. Figures and Legends……….….………...…………………………….56
Chapter III: CDK2/CyclinE-Mediated Phosphorylation on Threonine 199 of Nucleophosmin/B23 Localizes to Nuclear Speckles and Represses Pre-mRNA Splicing……………………………………...……………...70 A. Abstract……………………………………………………………….71 B. Introduction…………………………………………………………...72 C. Materials and Methods………………………………………………..74 D. Results ………………………………………………………………...79 E. Discussion…………………………………………………………….88 F. References…………………………………………………………….92 G. Figures and Legends………….……………………………………….97
Chapter IV: Ubiquitination of Nucleosphomin/B23………………….……………....107 A. Abstract……………………………………………...………………108 B. Introduction………………………………………………………….109 C. Materials and Methods………………………………………………112
1 D. Results……………………………………………………………….114 E. Discussion…………………………………………………………...117 F. References…………………………………………………………...119 G. Figures and Legends……………………….……………...………....122
Chapter V: Summary and Conclusions……………………………………………….128 A. Summary………………………………………………....…………..129 B. NPM/B23 and centrosome duplication………………………………129 C. NPM/B23 at the nuclear speckles……………………………………133 D. References……………………………………………………………137
Chapter VI: Future Directions………………………………………………………...140
LIST OF TABLES AND FIGURES
Chapter I: Introduction Figure 1. The structure of NPM/B23………………………………………………….8 Figure 2. Schematic diagram of the centrosome……………………………………...11 Figure 3. The centrosome (centriole) duplication cycle………………………………13 Figure 4. Schematic diagram of nuclear speckles…………………………………….16 Figure 5. Spliceosome complex formation along splicing reaction…………………..19
Chapter II: Specific Phosphorylation of Nucleophosmin on Thr199 by Cyclin- dependent Kinase 2-Cyclin E and Its Role in Centrosome Duplication Figure 1. CDK2/cyclin E phosphorylates NPM/B23 in vitro on threonine residue(s)…………………………………………………………………...63 Figure 2. NPM/B23 is phosphorylated by CDK2/cyclin E on threonine 199 (Thr199) in vitro……………………………………………………………64 Figure 3. The site of NPM/B23 phosphorylated in vitro by CDK2/cyclin E is phosphorylated in vivo……………………………………………………...65 Figure 4. Initiation of centrosome duplication is blocked by NPM/T199A expression…………………………………….…………………………….66 Figure 5. Expression of NPM/T199A mutant specifically inhibits centrosome duplication…………………………………………….……………………67 Figure 6. Expression of NPM/T199A mutant results in a high frequency of monopolar mitosis………………………………………………………….68 Figure 7. NPM/T199A physically associates with centrosomes……………………...69
Chapter III: CDK2/CyclinE-Mediated Phosphorylation on Threonine 199 of Nucleophosmin/B23 Localizes to Nuclear Speckles and Represses Pre-mRNA Splicing Figure 1. Phospho-Thr199 NPM antibody specifically recognizes NPM/B23 phosphorylated on Thr199 by CDK2/Cyclin E..…………………………..101
2 Figure 2. Cell-cycle dependent changes in the level of phospho-Thr199 NPM……….102 Figure 3. Phospho-Thr199 NPM co-localizes with splicing factors and re-distributes upon transcription inhibition………………………………...103 Figure 4. Phosphorylation on Thr199 enhances the RNA-binding activity of NPM/B23…………………………………………………………………..104 Figure 5. Phosphorylation on Thr 199 influences association of NPM/B23 with splicing factors……………………………………………………………..105 Figure 6. Phosphorylation on Thr 199 represses splicing in vitro……...…………….106
Chapter IV: CDK2/CyclinE Phosphorylation Dependent Ubiquitination of Nucleosphomin/B23 Figure 1. Proteasome inhibitor blocks centrosome duplication………………………124 Figure 2. Proteasome inhibition prevents loss of centrosomal NPM/B23……………125 Figure 3. NPM/B23 is ubiquitinated during G1 phase in vivo………………………..126 Figure 4. In vitro ubiquitination of NPM/B23………………………………………..127
Chapter V: Summary and Conclusion Figure 1. 3D-deconvolution analysis of NPM/B23 at the centrosome……………….131
3
Chapter I
Introduction
4 NPM/B23 as a multi-functional protein
Nucleophosmin (NPM/B23), also known as numatrin or NO38, is a
predominantly nucleolar phosphoprotein that has been implicated in many cellular
functions. NPM/B23 was initially identified as a protein that localized to the granular
and fibrillar regions of the nucleolus (1), and was found to associate with pre-ribosomal
particles (2). NPM/B23 is thought to function in ribosomal biogenesis, which is the
synthesis and processing of precursor ribosomal RNA (pre-rRNA) and the assembly of
ribosomal proteins on rRNA to form premature ribosomes. Ribosomal proteins are
synthesized in the cytoplasm, imported into the nucleoli for ribosomal assembly, and the
mature ribosome is then exported to the nucleus and cytoplasm (3). NPM/B23 is thought
to function in the intranuclear transport of pre-ribosomal particles (1,4-6). In addition,
NPM/B23 has an intrinsic endoribonuclease activity that cleaves within the internal
spacer sequences (ITS2 region) of the pre-rRNA, which is a crucial step in the process of rRNA maturation (7,8).
NPM/B23 has also been shown to play a role in nucleocytoplasmic transport (5).
This was shown by monitoring migration of nuclear proteins in interspecies heterokaryons and also by monitoring the distribution of antibodies injected into the cytoplasm of the cells. NPM/B23 has been implicated in nuclear import because it binds to the nuclear localization signal (NLS) of proteins such as Rex protein, a post- transcriptional regulator of human T-cell leukemia virus type I (9), Rev and Tat proteins of human immunodeficiency virus (10-13), SV40 T-antigen (14), and nucleolar p120 protein (15).
5 NPM/B23 has been shown to have molecular chaperone activity. The notion that
NPM/B23 may function as a molecular chaperone arose from the observation that
NPM/B23 stimulates nuclear import of the HIV protein, Rev. Under normal conditions
Rev had a tendency to aggregate (16) but this aggregation was reduced in the presence of
NPM/B23. It was thus proposed that NPM/B23 may act as a molecular chaperone to
prevent Rev aggregation. Indeed, NPM/B23 was found to inhibit Rev protein aggregation during thermal denaturation. Also, NPM/B23 was shown to promote the renaturation of denatured enzymes, thus restoring their activity (17). NPM/B23 was shown to preferentially bind to denatured proteins, exposing the hydrophobic regions in the NPM-substrate protein complex, and thus reducing protein aggregation. Recently,
NPM/B23 was found to bind to histones and mediate nucleosome assembly, thus exhibiting histone chaperoning activity (18).
Finally, NPM/B23 has been implicated in the regulation of cellular proliferation.
Over-expression of NPM/B23 results in the malignant transformation of cells (19).
NPM/B23 inhibits DNA-binding and transcription activity of interferon regulator factor-1
(IRF-1), a tumor suppressor (20). Moreover, NPM/B23 binds to retinoblastoma (Rb) protein and synergistically stimulates DNA polymerase α, which is involved in the initiation step of DNA replication (21). A role for NPM/B23 in cellular resistance to UV- irradiation has also been shown. Expression levels of NPM/B23 are lower in UV- sensitive cells (22) increase after UV irradiation (23), resulting in cellular resistance to
UV-induced growth inhibition and cell death (24,25). The mechanism by which this occurs is not yet fully understood. More recently, however, NPM/B23 was found to directly bind to the tumor suppressor p53 under cellular stress conditions (UV-
6 irradiation), thus stabilizing and increasing the transcriptional activity of p53 (as measured by increased expression of p21). This interaction could cause the observed cell cycle arrest or apoptosis (26). Recently, Kurki et al. have shown that UV-irradiation causes NPM/B23 to interact with MDM2. MDM2 normally binds to and inhibits p53, but upon UV-irradiation, NPM/B23 sequesters MDM2 away from p53, thus allowing for p53 stabilization and activity (27).
NPM/B23 has been shown to be more abundant in tumor and growing cells than in normal and resting cells (28-30). NPM/B23 is involved in several tumor related chromosome translocations such as anaplastic lymphoma kinase (NPM-ALK) (31), myeloid leukemia factor 1 (NPM-MLF1) (32) and retinoic acid receptor (NPM-RAR)
(33,34). Upon translocation, the oligomerization domain in the N-terminal portion of
NPM/B23 juxtaposes the catalytic domain of the various NPM/B23 partner proteins, thus
NPM/B23 is not thought to play a role in their transforming ability but instead provides a dimerization domain for their oligomerization.
NPM/B23 functional domains
NPM/B23’s function in diverse cellular events is mediated by its ability to bind to different proteins (Figure 1). The N-terminal homodimerization domain (residues 1-118) has been shown to be important for NPM/B23’s oligomerization (4,35). This N-terminal oligomerization domain and the first acidic domain are both essential for NPM/B23’s molecular chaperone activity (36). The heterodimerization domain (residues 186-260) has been shown to interact with many different proteins, such as p53, p120, and the HIV protein, Tat and is critical to NPM/B23’s ability to function in nucleocytoplasmic
7 shuttling. The C-terminal region has an aromatic ring domain and was found to be essential for NPM/B23’s DNA and RNA binding and endoribonuclease activities (37).
Moreover, it influences NPM/B23’s subcellular localization. NPM/B23 has been shown to exist in two splice variants, NPM/B23.1 (full length isoform) and NPM/B23.2 (missing
C-terminal 35 amino acids). NPM/B23.1 localizes mainly in the nucleoli, whereas
NPM/B23.2 is found primarily in the cytoplasm and at low levels in the nucleoplasm
(38). The difference in the localization pattern of the two isoforms was found to be due to the absence of 35 amino acids in the C-terminus of NPM/B23.2. The function of
NPM/B23.2 is not known. NPM/B23 discussed in this dissertation is the full length
NPM/B23.1 isoform.
P P P P P
1 118 186 260 294 a.a.
Homodimerization Domain Heterodimerization Domain Oligomerization p53, Tat binding domain
Molecular chaperone activity B23.1 unique domain DNA and RNA binding domain Ribonuclease activity
Figure 1. The structure of NPM/B23.
NPM/B23 consists of 294 amino acids and can be subdivided into three
major domains. The homodimerzation domain ( ) is necessary for
8 oligomerization and molecular chaperone activity. The heterodimerization
domain ( ) binds to p53, 120, Tat protein of HIV. The aromatic ring
domain ( ), is unique to the NPM/B23.1 isoform and contains the DNA
and RNA binding region, is important for NPM/B23’s ribonuclease
activity. (Acidic regions ( ), nuclear localization signal
( ) and phosphorylation (Ê)).
NPM/B23 function is mediated by phosphorylation events
NPM/B23 is a phosphoprotein and its phosphorylation plays a critical role in determining NPM/B23’s function. NPM/B23 is phosphorylated by several different kinases, including those that are expressed throughout the cell cycle, nuclear kinase II
(NKII) (39) and casein kinase II (CKII) (40), as well as by those whose expression varies with the cell cycle, such as Cyclin Dependent Kinase 1 (CDK1) (41), CDK2 (42) and
Polo-like Kinase 1 (Plk1) (43).
Phosphorylation of NPM/B23 by CKII has been shown to be important for
NPM/B23’s nucleocytoplasmic shuttling activity. For example, it has been shown that this phosphorylation event enhances NPM/B23’s affinity for peptides containing the NLS sequence (14) and also stimulates the rate of nuclear import of NLS-containing substrates
(10). Phosphorylation by CKII is also vital for NPM/B23’s molecular chaperone activity
(44). CKII phosphorylates NPM/B23 on serine 125, which is located within the acidic region of NPM/B23 that is essential for chaperone activity (36). Like other molecular chaperones, NPM/B23 binds to denatured substrates. The phosphorylation of NPM/B23 by CKII results in the dissociation of these substrates from NPM/B23.
9 NPM/B23 is phosphorylated by CDK1 kinase during mitosis (41). This phosphorylation event represses the RNA binding activity of NPM/B23 and is thought to result in the loss of NPM/B23’s ribosome biogenesis activity during mitosis (45).
Recently, another mitotic kinase, Plk1 was found to phosphorylate NPM/B23 on Serine 4
(43). Expression of the non-phosphorylatable NPM/B23-Ser 4 mutant resulted in abnormalities in centrosome duplication, chromosome segregation, and cytokinesis.
These effects are similar to those seen in the depleted Plk1 phenotype (46). This result implicates NPM/B23 as the downstream target of Plk1 during mitosis. NPM/B23 has also been implicated in the regulation of centrosome duplication. Our lab has found
NPM/B23 is present at the centrosome during G1 phase and is released just before the initiation of centrosome duplication occurs (42). My thesis examines the role of
CDK2/Cyclin E in regulating this function through NPM/B23, and will be discussed further in Chapter II.
Centrosomes and the Cell Cycle
The centrosome is a non-membranous organelle that is located close to the nucleus and is critical for proper cell division. It consists of a pair of centrioles that are surrounded by numerous protein aggregates that form the pericentriolar material (PCM).
Each centriole is made up of nine sets of triplet microtubules, and the centrioles are positioned perpendicular to each other (Figure 2).
10
Centriole pair
Pericentriolar material
Figure 2. Schematic diagram of the centrosome.
A pair of centrioles is surround by numerous aggregates of proteins known
as the pericentriolar material (PCM). Each centriole is cylindrical-shaped
and is composed of nine sets of triplet microtubules (~ 0.2 µm in diameter
and 0.2-0.5 µm in length).
The centrosome plays an important role in regulating cell cycle progression during interphase and mitosis. During interphase, the centrosome functions as a microtubule organization center (MTOC), in the nucleation of microtubules, as well as in establishing cell polarity and microtubule orientation (47). During mitosis, the centrosome directs the assembly of the bipolar spindles, which is necessary for proper chromosome segregation to each of the two daughter cells. In addition, the centrosome is
11 required for cell cycle progression as well as in the completion of cytokinesis (48-50).
Upon cytokinesis, each daughter cell inherits one centrosome and thus the centrosome must duplicate once prior to the next mitosis. This centrosome doubling mechanism occurs during interphase and is known as centrosome duplication. The centrosome duplication process in mammalian cells involves several steps (Figure 3). First, in late
G1 phase, loss of the orthogonal relationship and physical separation of the paired centrioles (mother and daughter centriole) occurs. Then, at the beginning of S phase, centriole duplication is initiated and a procentriole is synthesized at a right angle to each of the pre-existing mother and daughter centrioles. Lastly, the procentrioles elongate and mature by recruitment of pericentriolar material during S and G2 phase. The duplicated centrosomes migrate around the nucleus and are positioned orthogonally by the onset of
M phase (51).
Since centrosomes must duplicate once and only once during the cell cycle, the centrosome duplication cycle is tightly regulated and occurs in coordination with other cell cycle events (i.e. DNA synthesis). Disruption of the mechanisms that regulate the coordination of centrosome duplication with other cell cycle events leads to the abnormal amplification of centrosomes. Cells that have undergone centrosome amplification contain multiple centrosomes and generate defective spindle poles, which can result in chromosome segregation errors (52-54). Centrosome amplification is found in many human cancers and is thought to cause genomic instability, a hallmark of tumors (55-57).
Understanding how centrosome duplication is regulated has thus become essential to the field of cancer biology.
12 M
G1
G2
S
Figure 3. The centrosome (centriole) duplication cycle.
The initiation of centrosome duplication occurs at G1/S phases, and is followed by the elongation of the procentrioles during S and G2 phases.
During mitosis, centrosomes direct the bipolar spindle poles for proper
13 chromosome segregation, and by the end of mitosis, each daughter cell
will have inherited one centrosome.
Involvement of CDK2/Cyclin E in regulation of centrosome duplication
CDK2/Cyclin E has been shown to be responsible for cell cycle progression into
S phase (58). Expression of cyclin E occurs during mid-G1 phase and peaks in late G1
phase when DNA replication begins. DNA replication is initiated at the G1/S boundary
and is regulated by CDK2/Cyclin E. For example, CDK2/Cyclin E has been shown to
phosphorylate the retinoblastoma (Rb) protein, releasing bound E2F-transcriptional
factor, which stimulates transcription of genes involved in DNA replication (59).
Centrosome duplication occurs in coordination with DNA replication. Several
studies have implicated a role for CDK2/Cyclin E in the regulation of centrosome
duplication (60-62). For example, inhibition of p21cip1/waf1, a CDK2-inhibitor, in
mammalian cells has been shown to lead to centrosome amplification (63). Another
study used S-phase arrested Xenopus egg extracts, which were found to undergo multiple
cycles of centrosome duplication. However, when CDK2/Cyclin E activity was inhibited
in these extracts, centrosome duplication was inhibited; and when CDK2/Cyclin E was
added back to the extracts, centrosome duplication was restored (60). A third study
involved the use of S-phase arrested Chinese hamster ovary (CHO) cells, in which the
centrosome also duplicates multiple times. The addition of CDK2 inhibitor drugs or the
over-expression of p21cip1/waf again inhibited centrosome re-duplication (62). NPM/B23 was identified as one CDK2/Cyclin E centrosomal target (42). Using isolated centrosomes as substrates, Okuda et al. found that NPM/B23 is phosphorylated by
14 CDK2/Cyclin E and that this phosphorylation is important for the initiation of centrosome duplication (42), as further discussed in Chapter II. Another study has shown that CDK2/Cyclin E also indirectly regulates centrosome duplication by stabilizing Mps1 protein levels during S phase. Mps1 was originally identified as an essential protein kinase involved in the duplication of spindle pole body (the yeast equivalent of the centrosome) (64). Mps1 localizes to the centrosome throughout the cell cycle. In S phase arrested cells, which normally complete a single round of centrosome duplication, over-expression of Mps1 causes centrosomes to re-duplicate while over-expression of a kinase dead Mps1 mutant (Mps1-KD) blocks the centrosome duplication process (65,66).
CDK2/Cyclin E thus regulates centrosome duplication by targeting at least two centrosomal proteins, NPM/B23 and Mps1.
Nuclear speckles
The nucleus was one of the first intracellular structures identified under the microscope. Initially it was thought that little organization existed within the nuclear compartment; however, it has now become evident that the nucleus contains numerous specialized domains, some of which are extremely dynamic (67,68). For example, many nuclear factors are localized in distinct domains such as the nucleolus, the cajal body, the promyelocytic leukemia protein (PML) body and nuclear speckles.
Nuclear speckles were named by J. S. Beck in 1961 after its speckled-like pattern observed in rat-live sections immunolabelled with Sm autoimmune sera. They are enriched in pre-messenger RNA (pre-mRNA) splicing machinery, including the small nuclear ribonucleoprotein particles (snRNPs), spliceosome subunits and other non-
15 snRNP protein splicing factors. Nuclear speckles appear as punctate dots and surrounding diffuse areas distributed throughout the nucleoplasm, corresponding to interchromatin granule clusters (IGCs) and perichromatin fibrils (PFs), respectively
(Figure 4).
Figure 4. Schematic diagram of nuclear speckles.
Nuclear speckles consist of pre-mRNA splicing factors. The splicing
factors are localized in interchromatin granule clusters (IGC) and
perichromatin fibrils (PF). The splicing factors are thought to shuttle
between storage sites (IGC) and sites of active transcription (PF) (69).
The larger speckles (IGCs) are thought to function as storage/modification compartments that supply splicing factors to active sites of transcription (PFs), observed in the periphery of IGCs. There are several lines of evidence that support this view. For example, after cells are subjected to [3H] pulses, PFs are heavily labeled whereas IGCs are not labeled
16 (70). Moreover, in live cell studies, when cells were transfected with splicing factors fused to green fluorescent protein (GFP), the splicing factors were observed moving from speckles to the sites of transcription (71). Finally, inhibition of transcription or pre- mRNA splicing was found to result in an accumulation of splicing factors in enlarged, rounded speckles (IGCs) (72-74).
The mechanism that targets splicing factors to nuclear speckles is not fully understood. For some members of a serine/arginine (SR) protein family of non-snRNP splicing factors, the arginine/serine (RS) domain is thought to be the speckle-targeting signal (75). Phosphorylation within the RS domain of SR proteins has been shown to recruit SR splicing factors from IGCs to sites of active transcription (PFs) (76), and dephosphorylation is thought to shuttle the SR proteins back to the IGCs (77). The modulation of phosphorylation/dephosphorylation has also been implicated in the function of splicing factors in pre-mRNA splicing. It has been shown that the phosphorylation of some SR proteins and snRNP proteins is essential for the formation of functional spliceosomes (78,79). Also, the dephosphorylation event is necessary for completion of the splicing reaction and the disassembly of the spliceosome (80). Thus, assembly of the spliceosome is coordinated with pre-mRNA splicing through phosphorylation/dephosphorylation signaling pathways and by the spatial separation of inactive splicing factors in IGCs and active splicing factors in PFs.
Pre-mRNA splicing
The splicing of pre-mRNA is a crucial step in gene expression. The precise removal of pre-mRNA introns from nascent transcripts is essential for coding sequences
17 (exons) to be in frame for proper protein translation. Also, alternative splicing
mechanisms allow for many genes to generate functionally diverse protein isoforms,
expanding the coding potential of individual genes. Splicing is carried out by the
spliceosome, which consists of five uridine-rich small nuclear ribonucleoprotein (snRNP) complexes (U1, U2, U4, U5 and U6) as well as numerous non-snRNPs. For efficient splicing to occur, introns require a 5’ splice site, a branch point sequence and a 3’ splice site (Figure 5A). Splicing occurs in two steps: 1) 5’ splice site cleavage and lariat formation, 2) 3’ splice site cleave and exon ligation (81). The spliceosome assembles in vitro in an ordered fashion as shown in Figure 5B. Following the completion of splicing, the spliceosome is thought to disassemble and its components are recycled for another splicing cycle. However, there is now increasing evidence that preformed U1-U2-
U4/U6-U5 penta-snRNP complexes bind to pre-mRNA instead of forming step-by-step on the RNA, as was conventionally thought.
Moreover, recent proteomic analysis of spliceosome complexes has revealed that the spliceosome could be composed of over 300 proteins (review in (81,82)). Other than snRNPs, many non-snRNP proteins have been identified within the spliceosome and have been shown to be essential for the splicing reactions. For example, members of the SR family localize at the nuclear speckles and are thought to play a role in protein-protein interactions during spliceosome assembly. It is thus becoming increasingly evident that the spliceosome is a dynamic complex within which many of the proteins are not stably associated, and that the spliceosome composition changes during its assembly processes as well as during its catalytic activities on pre-mRNA.
18
A 5’ splice site 3’ splice site
Pre-mRNA Exon GU A AG Exon
Branch point
B U2 E Exon U1 A AG Exon
Exon U1 A Exon U2 U6
U5 U4
U1 Exon U6 B U5 U4 Exon U2
U1
Ex U4 on U6 C U5 U2 Exon
U6 U5 Exon Exon U2 + mRNA lariat intron
19 Figure 5. Spliceosome complex formation along splicing reaction.
A) Schematic diagram of pre-mRNA. B) Spliceosome complex
formation. First, complex E (early or commitment complex) is formed by
U1 snRNP binding at the 5’ splice site and loose binding of U2 snRNP to
the pre-mRNA. Complex A promotes a strong binding of U2 snRNP with
the branch point sequence. Next, complex B is formed by the association
of tri-snRNP (U5, U4/U6) and the exchange of U1 snRNP for U6 snRNP
binding at the 5’ splice site. Following the dissociation of U4 snRNP,
catalytically active C complex is formed and splicing occurs.
Ubiquitination
Ubiquitination is a post-translational modifications that is best known for tagging proteins for degradation (83). The ubiquitin moiety attaches to the target protein (or another ubiquitin moiety) through a lysine side chain. Ubiquitination occurs in a sequential manner and involves three classes of enzymes: E1 (ubiquitin activating enzyme), E2 (ubiquitin conjugating enzyme) and E3 (ubiquitin protein ligase). First, E1 activates the ubiquitin by the formation of a thiol-ester bond and then the active ubiquitin is transferred from E1 onto E2. E2 then binds to E3 and ubiquitin is transferred to a lysine residue on the substrate protein (84). Substrate specificity is conferred through E3.
To date, four substrate-specific E3 motifs have been identified: HECT (Homologous to
E6-AP Carboxyl Terminus) domain, RING (Really Interesting New Gene) finger domain,
U-box domain, and PHD domain (Plant Homeo-Domain) (85-87). Because most proteins can be ubiquitinated, hundreds of E3 ligases have been predicted by database analysis.
20 Recently, it has been shown that different types of ubiquitin modifications can occur to regulate protein function towards pathways other than protein degradation.
There are five lysine residues on ubiquitin (Lys6, Lys11, Lys29, Lys48, Lys63) that are involved in forming mono- or poly-ubiquitin chains (88-92). Conjugation of a single ubiquitin moiety (mono-ubiqutination) through Lys6 can target the substrate to function in transcription, histone function, endocytosis, or membrane trafficking (93). Attachment of Lys63-linked poly-ubiquitin is thought to modulate protein function through conformational changes, protein-protein interactions and changes in subcellular localization (94,95). Both Lys6- and Lys63-linked ubiquitination stabilize their protein substrates. However, Lys48-, Lys11- and Lys29-linked poly-ubiquitin chains are thought to target proteins to proteasome-mediated degradation. The best characterized ubiquitin mechanism, Lys48-linked poly-ubiquitination, regulates diverse processes such as cell cycle progression, differentiation, development, transcription, and signal transduction by the rapid and irreversible degradation of key regulatory proteins (96).
21 References
1. Spector, D. L., Ochs, R. L., and Busch, H. (1984) Chromosoma 90, 139-148
2. Prestayko, A. W., Klomp, G. R., Schmoll, D. J., and Busch, H. (1974)
Biochemistry 13, 1945-1951
3. Olson, M. O., Hingorani, K., and Szebeni, A. (2002) Int Rev Cytol 219, 199-266
4. Schmidt-Zachmann, M. S., Hugle-Dorr, B., and Franke, W. W. (1987) Embo J 6,
1881-1890
5. Borer, R. A., Lehner, C. F., Eppenberger, H. M., and Nigg, E. A. (1989) Cell 56,
379-390
6. Yung, B. Y., Busch, H., and Chan, P. K. (1985) Biochim Biophys Acta 826, 167-
173
7. Herrera, J. E., Savkur, R., and Olson, M. O. (1995) Nucleic Acids Res 23, 3974-
3979
8. Savkur, R. S., and Olson, M. O. (1998) Nucleic Acids Res 26, 4508-4515
9. Adachi, Y., Copeland, T. D., Hatanaka, M., and Oroszlan, S. (1993) J Biol Chem
268, 13930-13934
10. Szebeni, A., Mehrotra, B., Baumann, A., Adam, S. A., Wingfield, P. T., and
Olson, M. O. (1997) Biochemistry 36, 3941-3949
11. Li, Y. P. (1997) J Virol 71, 4098-4102
12. Marasco, W. A., Szilvay, A. M., Kalland, K. H., Helland, D. G., Reyes, H. M.,
and Walter, R. J. (1994) Arch Virol 139, 133-154
13. Fankhauser, C., Izaurralde, E., Adachi, Y., Wingfield, P., and Laemmli, U. K.
(1991) Mol Cell Biol 11, 2567-2575
22 14. Szebeni, A., Herrera, J. E., and Olson, M. O. (1995) Biochemistry 34, 8037-8042
15. Valdez, B. C., Perlaky, L., Henning, D., Saijo, Y., Chan, P. K., and Busch, H.
(1994) J Biol Chem 269, 23776-23783
16. Wingfield, P. T., Stahl, S. J., Payton, M. A., Venkatesan, S., Misra, M., and
Steven, A. C. (1991) Biochemistry 30, 7527-7534
17. Szebeni, A., and Olson, M. O. (1999) Protein Sci 8, 905-912
18. Okuwaki, M., Matsumoto, K., Tsujimoto, M., and Nagata, K. (2001) FEBS Lett
506, 272-276
19. Kondo, T., Minamino, N., Nagamura-Inoue, T., Matsumoto, M., Taniguchi, T.,
and Tanaka, N. (1997) Oncogene 15, 1275-1281
20. Tanaka, N., Ishihara, M., Kitagawa, M., Harada, H., Kimura, T., Matsuyama, T.,
Lamphier, M. S., Aizawa, S., Mak, T. W., and Taniguchi, T. (1994) Cell 77, 829-
839
21. Takemura, M., Sato, K., Nishio, M., Akiyama, T., Umekawa, H., and Yoshida, S.
(1999) J Biochem (Tokyo) 125, 904-909
22. Hirano, J., Wang, X., Kita, K., Higuchi, Y., Nakanishi, H., Uzawa, K., Yokoe, H.,
Tanzawa, H., Yamaura, A., Yamamori, H., Nakajima, N., Nishikiori, C., and
Suzuki, N. (2000) Cancer Lett 153, 183-188
23. Wu, M. H., Chang, J. H., Chou, C. C., and Yung, B. Y. (2002) Int J Cancer 97,
297-305
24. Higuchi, Y., Kita, K., Nakanishi, H., Wang, X. L., Sugaya, S., Tanzawa, H.,
Yamamori, H., Sugita, K., Yamaura, A., and Suzuki, N. (1998) Biochem Biophys
Res Commun 248, 597-602
23 25. Wu, M. H., Chang, J. H., and Yung, B. Y. (2002) Carcinogenesis 23, 93-100
26. Colombo, E., Marine, J. C., Danovi, D., Falini, B., and Pelicci, P. G. (2002) Nat
Cell Biol 4, 529-533
27. Kurki, S., Peltonen, K., Latonen, L., Kiviharju, T. M., Ojala, P. M., Meek, D., and
Laiho, M. (2004) Cancer Cell 5, 465-475
28. Chan, W. Y., Liu, Q. R., Borjigin, J., Busch, H., Rennert, O. M., Tease, L. A., and
Chan, P. K. (1989) Biochemistry 28, 1033-1039
29. Schmidt-Zachmann, M. S., and Franke, W. W. (1988) Chromosoma 96, 417-426
30. Feuerstein, N., Chan, P. K., and Mond, J. J. (1988) J Biol Chem 263, 10608-
10612
31. Morris, S. W., Kirstein, M. N., Valentine, M. B., Dittmer, K. G., Shapiro, D. N.,
Saltman, D. L., and Look, A. T. (1994) Science 263, 1281-1284
32. Yoneda-Kato, N., Look, A. T., Kirstein, M. N., Valentine, M. B., Raimondi, S. C.,
Cohen, K. J., Carroll, A. J., and Morris, S. W. (1996) Oncogene 12, 265-275
33. Redner, R. L., Rush, E. A., Faas, S., Rudert, W. A., and Corey, S. J. (1996) Blood
87, 882-886
34. Redner, R. L., Chen, J. D., Rush, E. A., Li, H., and Pollock, S. L. (2000) Blood
95, 2683-2690
35. Herrera, J. E., Correia, J. J., Jones, A. E., and Olson, M. O. (1996) Biochemistry
35, 2668-2673
36. Hingorani, K., Szebeni, A., and Olson, M. O. (2000) J Biol Chem 275, 24451-
24457
24 37. Wang, D., Baumann, A., Szebeni, A., and Olson, M. O. (1994) J Biol Chem 269,
30994-30998
38. Wang, D., Umekawa, H., and Olson, M. O. (1993) Cell Mol Biol Res 39, 33-42
39. Chan, P. K., Liu, Q. R., and Durban, E. (1990) Biochem J 270, 549-552
40. Chan, P. K., Aldrich, M., Cook, R. G., and Busch, H. (1986) J Biol Chem 261,
1868-1872
41. Peter, M., Nakagawa, J., Doree, M., Labbe, J. C., and Nigg, E. A. (1990) Cell 60,
791-801
42. Okuda, M., Horn, H. F., Tarapore, P., Tokuyama, Y., Smulian, A. G., Chan, P. K.,
Knudsen, E. S., Hofmann, I. A., Snyder, J. D., Bove, K. E., and Fukasawa, K.
(2000) Cell 103, 127-140
43. Zhang, H., Shi, X., Paddon, H., Hampong, M., Dai, W., and Pelech, S. (2004) J
Biol Chem
44. Szebeni, A., Hingorani, K., Negi, S., and Olson, M. O. (2003) J Biol Chem 278,
9107-9115
45. Okuwaki, M., Tsujimoto, M., and Nagata, K. (2002) Mol Biol Cell 13, 2016-2030
46. Liu, X., and Erikson, R. L. (2003) Proc Natl Acad Sci U S A 100, 5789-5794
47. Bornens, M. (2002) Curr Opin Cell Biol 14, 25-34
48. Piel, M., Nordberg, J., Euteneuer, U., and Bornens, M. (2001) Science 291, 1550-
1553
49. Khodjakov, A., and Rieder, C. L. (2001) J Cell Biol 153, 237-242
50. Hinchcliffe, E. H., Miller, F. J., Cham, M., Khodjakov, A., and Sluder, G. (2001)
Science 291, 1547-1550
25 51. Lange, B. M., and Gull, K. (1996) Trends Cell Biol 6, 348-352
52. Brinkley, B. R., and Goepfert, T. M. (1998) Cell Motil Cytoskeleton 41, 281-288
53. Tarapore, P., and Fukasawa, K. (2000) Cancer Invest 18, 148-155
54. Fukasawa, K., Choi, T., Kuriyama, R., Rulong, S., and Vande Woude, G. F.
(1996) Science 271, 1744-1747
55. Lingle, W. L., Lutz, W. H., Ingle, J. N., Maihle, N. J., and Salisbury, J. L. (1998)
Proc Natl Acad Sci U S A 95, 2950-2955
56. Lingle, W. L., and Salisbury, J. L. (2000) Curr Top Dev Biol 49, 313-329
57. Carroll, P. E., Okuda, M., Horn, H. F., Biddinger, P., Stambrook, P. J., Gleich, L.
L., Li, Y. Q., Tarapore, P., and Fukasawa, K. (1999) Oncogene 18, 1935-1944
58. Dulic, V., Lees, E., and Reed, S. I. (1992) Science 257, 1958-1961
59. Dyson, N. (1998) Genes Dev 12, 2245-2262
60. Hinchcliffe, E. H., Li, C., Thompson, E. A., Maller, J. L., and Sluder, G. (1999)
Science 283, 851-854
61. Lacey, K. R., Jackson, P. K., and Stearns, T. (1999) Proc Natl Acad Sci U S A 96,
2817-2822
62. Matsumoto, Y., Hayashi, K., and Nishida, E. (1999) Curr Biol 9, 429-432
63. Mantel, C., Braun, S. E., Reid, S., Henegariu, O., Liu, L., Hangoc, G., and
Broxmeyer, H. E. (1999) Blood 93, 1390-1398
64. Winey, M., Goetsch, L., Baum, P., and Byers, B. (1991) J Cell Biol 114, 745-754
65. Fisk, H. A., Mattison, C. P., and Winey, M. (2003) Proc Natl Acad Sci U S A 100,
14875-14880
66. Fisk, H. A., and Winey, M. (2001) Cell 106, 95-104
26 67. Spector, D. L. (1993) Curr Opin Cell Biol 5, 442-447
68. Misteli, T. (2001) Science 291, 843-847
69. Spector, D. L. (1996) Exp Cell Res 229, 189-197
70. Fakan, S. (1994) Trends Cell Biol 4, 86-90
71. Misteli, T., Caceres, J. F., and Spector, D. L. (1997) Nature 387, 523-527
72. Spector, D. L., Schrier, W. H., and Busch, H. (1983) Biol Cell 49, 1-10
73. O'Keefe, R. T., Mayeda, A., Sadowski, C. L., Krainer, A. R., and Spector, D. L.
(1994) J Cell Biol 124, 249-260
74. Huang, S., Mayeda, A., Krainer, A. R., and Spector, D. L. (1997) Mol Biol Cell 8,
1143-1157
75. Caceres, J. F., Misteli, T., Screaton, G. R., Spector, D. L., and Krainer, A. R.
(1997) J Cell Biol 138, 225-238
76. Misteli, T., Caceres, J. F., Clement, J. Q., Krainer, A. R., Wilkinson, M. F., and
Spector, D. L. (1998) J Cell Biol 143, 297-307
77. Misteli, T., and Spector, D. L. (1996) Mol Biol Cell 7, 1559-1572
78. Mermoud, J. E., Cohen, P. T., and Lamond, A. I. (1994) Embo J 13, 5679-5688
79. Roscigno, R. F., and Garcia-Blanco, M. A. (1995) Rna 1, 692-706
80. Mermoud, J. E., Cohen, P., and Lamond, A. I. (1992) Nucleic Acids Res 20, 5263-
5269
81. Jurica, M. S., and Moore, M. J. (2003) Mol Cell 12, 5-14
82. Nilsen, T. W. (2003) Bioessays 25, 1147-1149
83. Weissman, A. M. (2001) Nat Rev Mol Cell Biol 2, 169-178
84. Fang, S., and Weissman, A. M. (2004) Cell Mol Life Sci 61, 1546-1561
27 85. Huibregtse, J. M., Scheffner, M., Beaudenon, S., and Howley, P. M. (1995) Proc
Natl Acad Sci U S A 92, 2563-2567
86. Freemont, P. S. (2000) Curr Biol 10, R84-87
87. Patterson, C. (2002) Sci STKE 2002, PE4
88. Arnason, T., and Ellison, M. J. (1994) Mol Cell Biol 14, 7876-7883
89. Johnson, E. S., Ma, P. C., Ota, I. M., and Varshavsky, A. (1995) J Biol Chem 270,
17442-17456
90. Spence, J., Sadis, S., Haas, A. L., and Finley, D. (1995) Mol Cell Biol 15, 1265-
1273
91. Baboshina, O. V., and Haas, A. L. (1996) J Biol Chem 271, 2823-2831
92. Peng, J., Schwartz, D., Elias, J. E., Thoreen, C. C., Cheng, D., Marsischky, G.,
Roelofs, J., Finley, D., and Gygi, S. P. (2003) Nat Biotechnol 21, 921-926
93. Hicke, L. (2001) Nat Rev Mol Cell Biol 2, 195-201
94. Varadan, R., Assfalg, M., Haririnia, A., Raasi, S., Pickart, C., and Fushman, D.
(2004) J Biol Chem 279, 7055-7063
95. Passmore, L. A., and Barford, D. (2004) Biochem J 379, 513-525
96. Hershko, A., and Ciechanover, A. (1998) Annu Rev Biochem 67, 425-479
28
Chapter II
Specific Phosphorylation of Nucleophosmin on Thr199 by Cyclin- dependent Kinase 2-Cyclin E and Its Role in Centrosome Duplication
This chapter is a reprint of the material in the Journal of Biologial Chemistry (276: 21529-21537, 2001).
29 Abstract
The kinase activity of CDK2/cyclin E is required for centrosomes to initiate duplication.
We have recently found that nucleophosmin (NPM/B23), a phosphoprotein primarily
found in nucleolus, associates with unduplicated centrosomes, and is a direct substrate of
CDK2/cyclin E in centrosome duplication. Upon phosphorylation by CDK2/cyclin E,
NPM/B23 dissociates from centrosomes, which is a prerequisite step for centrosomes to
initiate duplication. Here, we identified that threonine 199 (Thr199) of NPM/B23 is the major phosphorylation target site of CDK2/cyclin E in vitro, and the same site is
phosphorylated in vivo. NPM/T199A, a non-phosphorylatable NPM/B23 substitution
mutant (Thr199 → Ala) acts as dominant-negative when expressed in cells, resulting in
specific inhibition of centrosome duplication. As expected, NPM/T199A remains
associated with the centrosomes. These observations provide direct evidence that the
CDK2/cyclin E-mediated phosphorylation on Thr199 determines association and
dissociation of NPM/B23 to the centrosomes, which is a critical control for the
centrosome to initiate duplication.
30 Introduction
The centrosome, a major microtubule organizing center of the animal cells, directs the formation of bipolar mitotic spindles, which is essential for accurate chromosome segregation to daughter cells (for reviews, see Refs. 1-3). Since each daughter cell inherits one centrosome upon cytokinesis, the centrosome must duplicate prior to the next mitosis, and do so only once. Thus, centrosome duplication must take place in coordination with other cell cycle events including DNA synthesis. In mammalian cells, the centriole, the core component of the centrosome, initiates duplication at the G1/S boundary (reviewed in Refs. 4-6). Activation of cyclin-dependent kinase 2 (CDK2)/cyclin E has recently been found to be essential for the centrosome to initiate duplication (7, 8). The activity of CDK2/cyclin E is regulated by the temporal expression of cyclin E, which normally occurs in late G1 (9, 10), and it has been known that active CDK2-cyclin E complexes is required for initiation of DNA replication (11,
12). These observations indicate that the late G1-specific activation of CDK2/cyclin E plays a key role for the coordinated initiation of centrosome and DNA duplication.
Indeed, we have shown that constitutive activation of CDK2/cyclin E by cyclin E overexpression in cultured mammalian cells results in uncoupling of the initiation of centrosome and DNA duplication: in these cells, the centrosomes initiate duplication in early G1 much before the onset of DNA synthesis (13). Unlike initiation of DNA synthesis which can only be triggered by CDK2-cyclin E after completion of a series of necessary events (14, 15), initiation of centrosome duplication appears to primarily depend on the activation of CDK2-cyclin E. Thus, the late G1-specific activation of
31 CDK2/cyclin E may serve as a checkpoint control for timely initiation of centrosome duplication.
We have recently identified nucleophosmin (NPM/B23) as a substrate of
CDK2/cyclin E in the initiation of centrosome duplication (16). NPM/B23, also called numatrin or NO38, was originally identified as a major nucleolar phosphoprotein localized in granular regions of the nucleolus, and has been shown to be associated with preribosomal particles (17-19). To date, NPM/B23 have been implicated in several distinct cellular functions, including assembly and/or intranuclear transport of preribosomal particles, cytoplasmic/nuclear trafficking, the regulation of DNA polymerase α activity, and centrosome duplication (16-21). NPM/B23 has also been shown to possess molecular chaperoning activities, including preventing protein aggregation, protecting enzymes during thermal denaturation, and facilitating renaturation of chemically denatured proteins (22). We have shown that NPM/B23 associates specifically with unduplicated centrosomes, and this association is controlled by CDK2/cyclin E-mediated phosphorylation, in which NPM/B23 loses its affinity to centrosomes in its phosphorylated form (16). Dissociation of the centrosomal NPM/B23 is essential for the centrosome to initiate duplication (16). For instance, microinjection of the anti-NPM/B23 monoclonal antibody, which blocks the CDK2/cyclin E-mediated phosphorylation of NPM/B23, inhibits centrosome duplication. Moreover, ectopic expression of this NPM/B23 deletion mutant (NPM∆186-239), which is unable to be phosphorylated by CDK2/cyclin E, results in suppression of centrosome duplication.
These results demonstrate that dissociation of centrosomal NPM/B23 by CDK2/cyclin E- mediated phosphorylation is critical for initiation of centrosome duplication, and that the
32 site(s) of NPM/B23 phosphorylated by CDK2/cyclin E lie within the sequence between
amino acid (a.a.) residue 186 and 239.
We here show that threonine residue 199 (Thr199) of NPM/B23 is specifically
phosphorylated by CDK2/cyclin E in vitro, and this phosphorylation is also observed in vivo. When NPM/B23 mutant with a substitution of this specific threonine residue to alanine (non-phosphorylatable) is expressed in cells, it affects on centrosome duplication in a dominant negative fashion, resulting in suppression of centrosome duplication.
These observations provide direct evidence that the CDK2/cyclin E-mediated phosphorylation of NPM/B23 on Thr199 is critical for dissociation of centrosomal
NPM/B23 and initiation of centrosome duplication.
33 Materials and Methods
Cells and transfection. Swiss 3T3 and HeLa cells were maintained in complete medium
[Dulbecco Modified Eagle Medium supplemented with 10% fetal bovine serum (FBS), penicillin (100 U/ml) and streptomycin (100 µg/ml)], in an atmosphere containing 10%
CO2.
For generation of HeLa cells overexpressing cyclin E, a plasmid encoding a
human cyclin E gene was co-transfected with a plasmid encoding a neomycin-resistant gene into HeLa cells by the calcium phosphate protocol. As a negative control, a vector was transfected. G418-resistant colonies arisen in the media containing G418 (800
µg/ml) at 2-3 weeks after transfection were sub-cloned, and analyzed for cyclin E expression. One cell line that overexpressed cyclin E (HeLa/CycE) and one vector- transfected G418-resistant cell line (HeLa/Vec) were maintained.
For transient transfection of wild-type and mutant NPM/B23 sequences, Swiss
3T3 cells were co-transfected with plasmids encoding either a FLAG-tagged wild-type or substitution mutant (Thr199 → Ala) NPM/B23 with a puromycin resistance gene plasmid
(pBabe/puro) at a molar ratio of 20:1 by the calcium phosphate protocol. After transfection in 37 °C for 8 h, cells were fed with fresh complete medium for 16 h. The cells were then treated with complete medium containing puromycin (4 µg/ml) for 36 h.
The puromycin resistant cells were pooled and replated on coverslips, and further cultured in fresh complete medium for 24 h.
Plasmid construction and purification of GST-NPM/B23. Mutant as well as wild-
type NPM/B23 cDNA sequences were fused to glutathione S-transferase using a two-step
PCR as described (23). The PCR-amplified products were inserted in frame into a
34 pGEX-4T-1 vector using BamH1 and EcoR1 restriction sites. GST-NPM fusion proteins
were bacterially purified according to the protocol provided by the manufacturer
(Amersham Pharmacia Biotech). Briefly, cells were induced with IPTG for 4 h before
harvesting. Clarified bacterial lysates were passed over a Sepharose 4B column, and
GST-NPM proteins were eluted. The concentration of the eluted GST-NPM was
estimated by comparison with bovine serum albumin with a known concentration run in
parallel in SDS-PAGE.
Immunoblot analysis. Cells were lysed in SDS/NP-40 lysis buffer [1% SDS, 1% NP-
40, 50 mM Tris (pH 8.0), 150 mM NaCl, 4 mM Pefabloc SC (Boerhinger Mannheim), 2
µg/ml leupeptin, 2 µg/ml aprotinin]. The lysates were boiled for 5 min, and then cleared by a 10 min centrifugation at 20,000 x g at 4oC. The supernatant was further denatured at
95oC for 5 min in sample buffer [2% SDS, 10% glycerol, 60 mM Tris (pH 6.8), 5% β-
mercaptoethanol, 0.01% bromophenol blue]. Samples were resolved by SDS-PAGE, and
transferred onto Immobilon-P (Millipore) sheets. The blots were first incubated in
blocking buffer [5% (wt/vol) nonfat dry milk in Tris-buffered saline (TBS) + Tween 20
(TBS-T)] for 1 h. The blots were then incubated with primary antibody for 2 h, followed
by incubation with horseradish peroxidase-conjugated secondary antibody for 1 h. All of
the procedures were performed at room temperature. The antibody-antigen complex was
visualized by ECL chemiluminescence (Amersham Pharmacia Biotech).
Indirect immunofluorescence. Cells grown on coverslips were fixed with 10%
formalin/10% methanol for 20 min at room temperature. The cells were permeabilized
with 1% NP-40 in phosphate-buffered saline (PBS) for 5 min, followed by incubation
with blocking solution (10% normal goat serum in PBS) for 1 h. Cells were then probed
35 with primary antibodies for 1 h, and antibody-antigen complexes were detected with
either rhodamine- or FITC-conjugated goat secondary antibody by incubation for 1 h at
room temperature. The samples were washed three times with PBS after each incubation,
and then counterstained with 4',6-diamidino-2-phenylindole (DAPI).
For co-immunostaining of α- and γ-tubulins to examine centriole pairs within
centrosomes, cells were first placed on ice for 30 min to destabilize microtubules
nucleated at the centrosomes. The cold-treated cells were then subjected to brief
extraction (∼30 sec) with cold extraction buffer [0.75% Triton X-100, 5mM Pipes, 2 mM
EGTA (pH 6.7)], briefly washed in cold PBS, and fixed with 10% formalin/10%
methanol. Cells were immunostained with anti-α-tubulin monoclonal (DM1A) and anti-
γ-tubulin polyclonal (24) antibodies. The antibody-antigen complexes were detected with
FITC-conjugated goat anti-rabbit IgG and rhodamine-conjugated goat anti-mouse IgG
antibodies.
In vitro Kinase Assay. For examination of CDK2/cyclin E activity in cyclin E-
overexpressing HeLa cells, cell lysates were subjected to immunoprecipitation using anti-
cyclin E antibody (Santa Cruz, sc-198). The antibody-antigen complexes were collected with protein A-agarose, and tested for a histone H1 kinase activity as described previously (13). For in vitro phosphorylation of NPM/B23 by CDK2/cyclin E,
CDK2/cyclin A or CDK1/cyclin B, GST, GST-NPM/wt, or GST-NPM/T199A were incubated with baculovirally purified active CDK2/cyclin E or CDK2/cyclin A complexes (25) or the immunoprecipitated CDK1/cyclin B from mitotically arrested
Swiss 3T3 cells by nocodazole-treatment using agarose-conjugated anti-cyclin B monoclonal antibody (GNS1, Santa Cruz) (26). The in vitro kinase reactions were
36 performed in 10 mM PIPES buffer in the presence of [γ-32P] ATP at 25°C for 15 min, and at 32°C for additional 15 min. The samples were resolved in SDS-PAGE, and the gel was dried and autoradiographed. For both histone H1 and GST-NPM kinase assays, 32P- incorporation was quantitated by scanning with Fuji Phosphoimager 1000.
BrdU incorporation assay. The assay was performed using the BrdU-labeling kit
(Boehringer Mannheim) according to the manufacturer’s instruction. Briefly, cells were
fixed in 70% ethanol in 50 mM glycine (pH 2.0) for 20 min at -20°C, incubated in the
blocking buffer for 1 h at room temperature, and then probed with anti-γ-tubulin polyclonal and anti-BrdU monoclonal antibodies for 30 min at 37°C. Antigen-antibody complexes were detected by FITC-conjugated sheep anti-mouse IgG and rhodamine- conjugated goat anti-rabbit IgG antibodies. Preparation of cells is described in the figure legend.
Phosphoamino acid analysis . In vitro phosphorylated GST-NPM/wt by CDK2/cyclinE in the presence of [32P-γ]ATP was resolved in SDS-PAGE. 32P-labeled NPM/B23 was
eluted from the gel and subjected to acid hydrolysis. The phosphorylated amino acids
were separated by two-dimensional electrophoresis on a thin layer cellulose gel plate as
described previously (27).
Two-dimensional (2D) tryptic phosphopeptide mapping. For preparation of in vivo
32P-labeled NPM/B23, HeLa/CycE cells were labeled for 2.5 h in phosphate free medium
containing 1 mCi/ml 32P-orthophosphate and 2% dialyzed FBS. Cells were lysed in lysis
buffer (1% Triton X-100, 50 mM Tris (pH 8.0), 50 mM β-glycerophosphate, 50 mM
sodium fluoride, 50 mM NaCl, 0.1% Sodium deoxycholate, 4 mM Pefabloc SC, 2 µg/ml
leupeptin, 2 µg/ml aprotinin, 4 µM Microcystin-LR). After pre-clearing with protein G-
37 conjugated agarose, the lysates were subjected to immunprecipitation with anti-NPM/B23 monoclonal antibody. Antibody-antigen complexes were collected by protein G- conjugated agarose, and were resolved by 10% SDS-PAGE. The gel was dried and autoradiographed. The in vivo 32P-labeled NPM/B23 proteins were eluted from the gel.
In vitro 32P-labeled GST-NPM/wt fusion proteins were prepared as described above for phosphoamino acid analysis. Two-dimensional tryptic phosphopeptide mapping was performed as described previously (27). Briefly, both in vitro labeled GST-NPM/wt and in vivo labeled NPM/B23 proteins were oxidized in performic acid, and digested with
TPCK-trypsin (Worthington). Eeach sample was loaded onto a thin layer cellulose gel plate, and run for 1h at 1200V at 4°C. The plates were dried, and subjected to chromatography (37.5% n-butanol, 25% pyridine, 7.5% acetic acid) in the vertical direction.
38 Results
Identification of the site of NPM/B23 specifically phosphorylated by CDK2/cyclin E
in vitro. We have previously shown that NPM/B23 is a direct centrosomal protein
substrate of a CDK2/cyclin E serine/threonine kinase complex in centrosome duplication.
NPM/B23 deletion mutant (∆186-239) fails to be phosphorylated by CDK2/cyclin E, and
acts as a dominant negative when expressed in cells (16), indicating that the CDK2/cyclin
E-mediated phosphorylation site(s) lie between a.a. 186 and 239. The sequence analysis
of human NPM/B23 revealed that there are several serine and threonine residues within
this region. By phosphoamino acid analysis, we first tested whether serine or threonine
residue(s) (or both) are phosphorylated by CDK2/cyclin E in vitro. The in vitro kinase
reaction of wild-type NPM/B23 fused to GST (GST-NPM/wt) was performed using
baculovirally purified active CDK2/cyclin E in the presence of [γ-32P]ATP. As negative
controls, GST proteins as well as GST-NPM(∆186-239) were used as substrates. The
kinase reaction samples were resolved in SDS-PAGE, and autoradiographed (Fig. 1A).
CDK2/cyclin E did not phosphorylate the GST moiety (lane 1). As shown previously,
GST-NPM(∆186-239) deletion mutant failed to be phosphorylated by CDK2/cyclin E
(lane 2), while GST-NPM(wt) was phosphorylated at a readily detectable level (lane 3).
The phosphorylated GST-NPM/wt proteins were eluted from the gel, and subjected to
phosphoamino acid analysis (Fig. 1B). We found that NPM/B23 was phosphorylated
exclusively on threonine residue(s) in vitro by CDK2/cyclin E.
There are four possible CDK2 phosphorylation consensus sequences within this region (Fig. 2A, indicated by arrowheads). To identify which threonine residue(s) are phosphorylated by CDK2/cyclin E, each of these four threonine residues (Thr199,
39 Thr219, Thr234, and Thr237) was replaced with alanine (non-phosphorylatable amino
acid). These mutants fused to GST (GST-NPM/T199A, GST-NPM/T219A, GST-
NPM/T234A and GST-NPM/T237A, respectively) were subjected to an in vitro kinase assay with CDK2/cyclin E (Fig. 2B, top panels). GST and GST-NPM/wt were included as controls in the experiment. Three GST-NPM mutants (T219A, T234A, and T237A) were phosphorylated (lanes 4-6) at similar levels with GST-NPM/wt (lane 2). However,
GST-NPM/T199A showed dramatically reduced phosphorylation (lane 3), suggesting that Thr199 is the primary phosphorylation site of NPM/B23 by CDK2/cyclin E in vitro.
CDK2 is known to be activated also by association with cyclin A, which is up- regulated during S and G2 phases (9, 10, 28-30). We, thus, examined whether
CDK2/cyclin A could phosphorylate Thr199. GST-NPM fusion proteins were subjected to an in vitro kinase assay with baculovirally purified CDK2/cyclin A (Fig. 2B, bottom panels). CDK2/cyclin A could phosphorylate all the mutants except GST-NPM/T199A at a similar efficiency with GST-NPM/wt, demonstrating that CDK2/cyclin A can also specifically phosphorylate Thr199 in vitro.
It has previously been shown that CDK1/cyclin B, a CDK/cyclin complex specifically activated during mitosis (reviewed in Ref. 31), phosphorylates NPM/B23
(32). However, the phosphorylation target site(s) of CDK1/cyclin B had not been identified. We, thus, tested whether CDK1/cyclin B phosphorylates the same threonine residue which is phosphorylated by CDK2/cyclin E. GST-NPM substitution mutants described above as well as GST-NPM/wt were subjected to an in vitro kinase assay with immuno-purified CDK1/cyclin B (Fig. 2C). CDK1/cyclin B phosphorylated GST-
NPM/T199A (lane 2) at a similar efficiency with GST-NPM/wt (lane 1), indicating that
40 Thr199 is not the target site of CDK1/cyclin B. In contrast, the levels of 32P-
incorporation of both GST-NPM/T234A (lane 4) and GST-NPM/T237A (lane 5) were
reduced to less than 50% of GST-NPM/wt. When both Thr234 and Thr237 were
replaced to alanine residues (GST-NPM/T234A/T237A), the level of 32P-incorporation became almost undetectable (lane 6). This result indicates that CDK1/cyclin B phosphorylates both Thr234 and Thr237 in vitro. Moreover, CDK2/cyclin E and
CDK1/cyclin B phosphorylate different sites of NPM/B23.
Thr199 of NPM/B23 is phosphorylated in vivo. We next examined whether Thr199 is
phosphorylated in vivo by 2D-tryptic peptide mapping of NPM/B23 prepared from
metabolically labeled cells with 32P-orthophosphate. Since CDK2/cyclin E is normally activated only in late G1, we assumed that CDK2/cyclin E-mediated phosphorylation of
NPM/B23 might not be efficiently detected if the exponentially growing cells are used.
In addition, it is not known whether CDK2/cyclin A-mediated phosphorylation of
NPM/B23 on Thr199 occurs in vivo in a similar manner as in vitro. To circumvent these problems, we first generated HeLa cells overexpressing cyclin E by transfecting human cyclin E together with a plasmid encoding neomycin-resistance gene as a selection
marker. It has been shown that overexpression of cyclin E results in constitutive
activation of CDK2/cyclin E (9, 13-15). The G418-resistant colonies were subcloned,
examined for cyclin E expression by immunoblot analysis, and one cell line that
overexpressed cyclin E (HeLa/Cyc E) was maintained for further experimentation (Fig.
3A). Consistent with the previous studies (9, 13-15), the immunoprecipitates from HeLa
/CycE cells using anti-cyclin E antibody showed a histone H1 kinase activity 4-5-fold
higher than the vector-transfected control HeLa cells (Hela/Vec) (Fig. 3B).
41 HeLa/CycE cells were metabolically labeled in the presence of 32P- orthophosphate. The cell lysates were immunoprecipitated with anti-NPM/B23 monoclonal antibody. The phospho-labeled NPM/B23 proteins were purified from the gel after fractionation of the immunoprecipitates in SDS-PAGE. The NPM/B23 proteins phosphorylated by CDK2/cyclin E in vitro were also purified after fractionation in SDS-
PAGE as described in the legend to Fig. 1. The purified in vivo- and in vitro-labeled
NPM/B23 were subjected to tryptic digestion, followed by a 2-D tryptic phosphopeptide
mapping analysis. In vitro phosphorylated NPM/B23 showed one specific spot (Fig. 3C,
panel a). In vivo phosphorylated NPM/B23 gave arise five major phosphopeptides (spots
1-5) (panel b). This is not unexpected, since NPM/B23 has been shown to be
phosphorylated by other kinases at several residues, including CDK1-cyclin B (32),
casein kinase II (33), and nuclear kinase (N-II) (34). However, one spot (spot 2 in panel
b, indicated by an arrow) showed a similar migration with that observed for the in vitro
phosphorylated NPM/B23. To confirm that these spots represented identical
phosphorylation sites, a mixture of the in vivo and in vitro labeled NPM/B23 was
analyzed. We found that these spots comigrated (panel c, indicated by an arrow),
demonstrating that the same tryptic peptide fragment was phosphorylated in vitro by
CDK2/cyclin E and in vivo in HeLa/CycE cells. When potential tryptic fragments were
deduced from the sequence, the primary fragment containing Thr199 was found to be
DTPAK, which contains only one threonine residue. Thus, we concluded that NPM/B23
is phosphorylated on Thr199 in vivo.
42 Suppression of centrosome duplication by expression of NPM/T199A. We have previously shown that the mutant NPM/B23 (NPM∆186-239) with a deletion of 54 amino acids, which includes the Thr199 CDK2/cyclin E phosphorylation site, acts as a dominant-negative when expressed in cells, resulting in inhibition of centrosome duplication (16). Although this observation strongly suggests that the dominant-negative activity of this mutant is attributed to not being able to be phosphorylated by
CDK2/cyclin E, it does not exclude the possibility that the deleted sequence other than the phosphorylation site may be also important for the regulation of centrosome duplication. We, thus, examined whether the non-phosphorylatable NPM/T199A mutant acts as a dominant-negative in centrosome duplication in a similar fashion with
NPM∆186-239. If expression of NPM/T199A mutant results in suppression of centrosome duplication, the phosphorylation of Thr199 is most likely a sole event necessary for the NPM/B23-dependent control of centrosome duplication. The FLAG- epitope tagged wild-type NPM/B23 (NPM/wt) and NPM/T199A were place in eukaryotic expression vectors, and transfected into Swiss 3T3 cells together with a plasmid encoding a puromycin-resistant gene. As a control, the vector was transfected. The puromycin- resistant cells selected by puromycin-treatment for 36 h were replated and cultured for additional 24 h. Cells were first examined for the level of expression of transfected
NPM/B23 by immunoblot analysis using anti-FLAG antibody (Fig. 4A). Both NPM/wt and NPM/T199A-transfectants expressed similar levels of transfected NPM/B23.
Cells were examined for centrosomes by immunostaining of γ-tubulin, a major component of pericentrial material of the centrosome (reviewed in Ref. 35). The graph in
Fig. 4B shows the centrosome profiles of the vector-, NPM/wt-, and NPM/T199A-
43 transfectants. In the vector-transfected cells, ∼40% of cells contained one centrosome, and ~60% contained two centrosomes. The cells transfected with NPM/wt showed similar centrosome profiles with the vector-transfectants. In contrast, the majority
(>80%) of cells transfected with NPM/T199A contained one centrosome, indicating that ectopic expression of NPM/T199A results in suppression of centrosome duplication.
To verify whether the anti-γ-tubulin antibody-reactive signals (dots) represented intact centrosomes with a pair of centrioles, cells were also immunostained for the centrioles. Since α-tubulin is one of the major constituents of centrioles, immunostaining of α-tubulin allows visualization of a centriole pair within the centrosome. Cells were subjected to cold treatment (which depolymerizes microtubules nucleated at the centrosomes), followed by a brief extraction prior to fixation (see Experimental
Procedures), and co-immunostained with anti-γ-tubulin polyclonal and anti-α-tubulin monoclonal antibodies (Fig. 4C). Each dot detected by anti-γ-tubulin antibody (panels a and e) was resolved to a pair of dots (representing a centriole pair) by anti-α-tubulin antibody at a higher magnification (panels b and f, panels I-VIII). All of the anti-γ- tubulin antibody reactive dots were co-immunostained by anti-α-tubulin antibody as doublets. Thus, the doublets detected by anti-γ-tubulin antibody represents duplicated centrosomes. The centrosome profiles determined by anti-α-tubulin antibody were similar to those determined by anti-γ-tubulin antibody (data not shown).
To eliminate the possibility that inhibition of centrosome duplication by
NPM/T199A is due to a general cell cycle arrest, Swiss 3T3 cells were transiently transfected with either a vector or a NPM/T199A mutant plasmid along with a plasmid encoding a puromycin resistance gene as a selection marker. The puromycin-resistant
44 cells selected during 36 h puromycin treatment were cultured for additional 24 h. During
the final 3 h of culturing, BrdU was added to the media to monitor cell cycling. Cells
were co-immunostained with anti-γ-tubulin polyclonal and anti-BrdU monoclonal
antibodies (Fig. 5A). Approximately 10% of vector-transfected and ∼6% of
NPM/T199A-transfected cells were BrdU-positive, suggesting that the expression of
NPM/T199A may be partially cytotoxic. Similar observation was previously made for
the NPM (∆186-239) deletion mutant (16). Examination of centrosomes revealed that all
of the BrdU-positive vector-transfected cells contained duplicated centrosomes, while the
majority (∼80%) of NPM/T199A-transfected BrdU-positive cells contained a single
centrosome (Fig. 5B & 5C). Thus, dominant negative activity of NPM/T199A
specifically targets the centrosome duplication process.
Aberrant mitoses with monopolar spindles resulting from expression of
NPM/T199A mutant. The finding, in which expression of NPM/T199A mutant results in suppression of centrosome duplication, but not DNA duplication, predicts that
NPM/T199A-transfected cells should progress through the cell cycle to mitosis without centrosome duplication. This should lead to mitosis with monopolar instead of bipolar spindles. To test this prediction, Swiss 3T3 cells were transiently transfected with either a vector or a NPM/T199A mutant plasmid along with a plasmid encoding a puromycin resistance gene as a selection marker. The puromycin-resistant cells were selected as described above (Fig. 4), and immunostained with anti-α- and β-tubulin monoclonal
antibodies and anti-γ-tubulin polyclonal antibody. The cells were also counterstained
with DAPI. Mitotic cells were identified by DAPI-stained condensed chromosomes, and
the number of spindle poles present in each mitotic cell was determined. Virtually all of
45 the vector-transfected mitotic cells contained two spindle poles (Fig. 6A), forming bipolar mitotic spindles (Fig. 6B, panels a-d). In contrast, ∼80% of the NPM/T199A- transfected mitotic cells contained single spindle poles (Fig. 6A) with disorganized microtubule staining (Fig. 6B, panels e-h). These observations further support the dominant-negative activity of the NPM/T199A mutant, specifically inhibiting duplication of centrosome.
NPM/T199A mutant associates with the centrosomes. We have previously shown that
NPM/B23 dissociates from centrosomes upon CDK2/cyclin E-mediated phosphorylation
(16), which implies that the NPM/T199A mutant should associate with centrosomes.
We, thus, examined the localization of transfected FLAG-epitope tagged NPM/T199A by co-immunostaining with anti-γ-tubulin polyclonal and anti-FLAG monoclonal antibodies
(Fig. 7). No anti-FLAG antibody staining was observed in the vector-transfected cells
(panel b). In contrast, anti-FLAG antibody detected a single dot adjacent to nucleus in the NPM/T199A-transfected cells (panel f), which overlap with the dot detected by anti-
γ-tubulin antibody (panels e and h). Thus, NPM/T199A mutant physically associates with centrosomes.
46 Discussion
Centrosome hyperamplification, which leads to formation of aberrant mitotic spindles, are now well accepted as one of major causes of chromosome instability in human cancers (36-39). In normal cells, the centrosome duplication cycle is tightly regulated. Coordinated initiation of centrosome and DNA duplication is one of the major regulatory checkpoint for proper progression of the centrosome duplication cycle, and it is established at least in part by the late G1-specific activation of CDK2/cyclin E (7, 8).
Activation of CDK2/cyclin E plays a major role in the initiation of DNA synthesis through phosphorylation of retinoblastoma susceptibility protein (pRb). When pRb is phosphorylated, it releases the pRb-bound E2F transcriptional factor, which then stimulates the transcription of a number of genes required for DNA synthesis (for reviews, see Refs. 40, 41). Requirement of E2F has also been implicated in the initiation of centrosome duplication (42), suggesting that E2F-dependent expression of specific protein(s) may be needed for centrosome duplication. Indeed, it has been shown that, in
Chinese hamster ovary cells, synthesis of certain centrosomal proteins during G1 is necessary for initiation of centrosome duplication (43). In addition, CDK2/cyclin E has been shown to directly act on a centrosomal protein to initiate centrosome duplication.
NPM/B23 binds specifically to unduplicated centrosomes, and loses its centrosome- binding activity when phosphorylated by CDK2/cyclin E. Dissociation of the centrosomal NPM/B23 appears to be pre-requisite for centrosomes to initiate duplication, since centrosome duplication is blocked by microinjection of anti-NPM/B23 antibody, which prevents CDK2/cyclin E-mediated phosphorylation and dissociation of centrosomal NPM/B23 (16).
47 We have previously shown that NPM/B23 deletion mutant [NPM(∆186-239)], which fails to be phosphorylated by CDK2/cyclin E, acts as a dominant-negative when expressed in cells, resulting in suppression of centrosome duplication (16). In this study, we identified Thr199 as the CDK2/cyclin E phosphorylation target site of NPM/B23 both in vitro and in vivo. The NPM/B23 mutant with an alanine substitution at this site
(NPM/T199A) acts as a dominant- negative when expressed in cells, and suppresses centrosome duplication, similar to NPM (∆186-239) deletion mutant. Initial stage of centrosome duplication consists of a series of distinct steps; loss of orthogonal configuration and physical separation of the centriole pair, which is followed by synthesis of a procentriole next to each preexisting centriole. Co-immunostaining of centrosomes in the NPM/T199A-transfected cells with anti-α-tubulin and anti-γ-tubulin antibodies detected anti-α-tubulin antibody-reactive doublets (a centriole pair) within the anti-γ- tubulin antibody-reactive dot (pericentriolar material), suggesting that the duplication of centrosomes in the NPM/T199A-transfected cells is blocked in the early stage. This is consistent with our earlier finding using cells expressing NPM(∆186-239), the non- phosphorylatable deletion mutant, by thin section transmission electron microscopy (16).
In these cells, unduplicated centrosomes are physically intact with the orthogonal configuration of the centriole pair, which is typical of those found in early G1-phase of the cell cycle. Thus, similar to the non-phosphorylatable deletion mutant, expression of
NPM/T199A mutant blocks the early step of centrosome duplication. Dominant-negative activity of NPM/T199A on centrosome duplication is further evidenced by the high frequency of aberrant mitoses with monopolar spindles in the NPM/T199A-transfected cells, resulting from cell cycle progression to mitosis, without centrosome duplication.
48 These observations provide direct evidence that CDK2/cyclin E-mediated phosphorylation of NPM/B23 comprises one of the key events in the initiation of centrosome duplication. At present, the molecular basis of the centrosome-binding property of NPM/B23 (i.e. which centrosomal protein(s) the unphosphorylated form of
NPM/B23 associates with) is unknown. The identification of the CDK2/cyclin E- mediated phosphorylation site will, however, expedite the elucidation of the particular centrosomal component(s) with which NPM/B23 directly associate.
Another important issue of the inhibition of centrosome duplication by expression of NPM/T199A mutant is the consequence of monopolar spindle formation. Considering the role of the centrosomes (spindle poles) in cytokinesis (47), it is safe to assume that monopolar mitotic cells do not undergo cytokinesis. If these cells enter the next cell cycle without cytokinesis, we should expect an increase in the number of cells with abnormal amplification of genome (≥ 8N). However, the flow cytometric analysis of the
NPM/T199A-transfected cells failed to detect any noticeable increase in the number of cells with abnormally amplified genome (data not shown), suggesting that formation of monopolar spindles likely leads to cell death. However, the mechanism of how cell death is induced in the monopolar mitotic cells remains to be clarified.
NPM/B23 associates with and dissociates from centrosomes in a cell cycle stage- specific manner (16, 44). During early to mid G1, NPM/B23 associates with the unduplicated centrosomes. In late G1, NPM/B23 dissociates from the centrosomes upon phosphorylation by CDK2/cyclin E. During S and G2 phases, association of NPM/B23 with the duplicated centrosomes is not detected. However, during mitosis, NPM/B23 re- associates with the centrosomes. This cell cycle stage-dependent dissociation and re-
49 association of NPM/B23 with centrosomes may be controlled by differential phosphorylation by CDK/cyclin complexes. CDK2/cyclin E activity peaks during late
G1, triggering dissociation of the centrosomal NPM/B23. Upon entry into S-phase, cyclin E expression becomes halted. Since cyclin E is intrinsically unstable, cyclin E- dependent CDK2 activity becomes minimal during S-phase (9, 10). In contrast, the level of cyclin A is low in late G1, but increases during S and G2 phases (27-30). Thus, during
S and G2 phases of the cell cycle, CDK2/cyclin A activity is high. We found that
CDK2/cyclin A could also phosphorylate NPM/B23 specifically on Thr199 in vitro at a similar efficiency with CDK2/cyclin E. Thus, it is possible that continual presence of active CDK2/cyclin A is responsible for preventing the re-association of NPM/B23 to centrosomes during S and G2. Moreover, NPM/B23 has previously been shown to be phosphorylated by CDK1/cyclin B, a mitotic CDK/cyclin complex (32). We found that
CDK1/cyclin B specifically phosphorylates Thr234 and Thr237 in vitro, which are different from CDK2/cyclin E (and cyclin A)-mediated phosphorylation site. It remains to be investigated whether phosphorylation of Thr234 and/or Thr237 by CDK1/cyclin B is required for re-association of NPM/B23 with the centrosomes during mitosis. These questions are currently addressed in our laboratory.
NPM/B23 has been shown to participate in various cellular events that are to all appearances unrelated to each other, including ribosome assembly, intracellular trafficking, DNA polymerase activity, and centrosome duplication. These diverse functions of NPM/B23 are perhaps attributed to its molecular chaperoning activity as reported previously (22). All the cellular events, in which NPM/B23 has been shown to function, involve either large multi-protein complexes or organelles consisting of many
50 different proteins in a crowded condition. Thus, association/dissociation of NPM/B23
may dramatically influence the centrosome proper, and thus determine the structural as
well as functional state of the centrosome. Moreover, the CDK2/cyclin E-mediated
functional modification of NPM/B23 may target other cellular event(s) as well, since the
BrdU-incoporation assay showed that expression of NPM/T199A mutant partially blocks
(or slows down) the cell cycle progression.
Acknowledgements – We thank Drs. J. Roberts for a human cyclin E plasmid, P. -K.
Chan for anti-NPM/B23 antibody and NPM/B23 plasmids, D. Morgan for baculovirus stocks for CDK2/cyclin A and CDK2/cyclin E. We also thank T. Kim for his technical assistance. This research is supported in part by Cancer Research Challenge and Ruth
Lyons Fund.
51 References
1. Brinkley, B. R. (1985) Ann. Rev. Cell Biol., 1, 145-172
2. Bornens, M. (1992) The Centrosome (V.I. Kalnins, ed.) Academic Press, New York,
pp. 1-43
3. Lange, B. M., and Gull, K. (1996) Trends Cell Biol., 6, 348-352
4. Mazia, D. (1987) Int. Rev. Cytology 100, 49-92
5. Vandre, D. D., and Borisy, G. G. (1989) Mitosis: Molecules and Mechanisms (J.S.
Hyams and B.R. Brinkley, eds.), Academic press, San Diego, pp. 39-75
6. Tournier, F., and Bornens, M. (1994) Microtubules (J.S. Hyams and C.W. Lloyd eds.)
Wiley-Liss, New York, pp. 303-324
7. Hinchcliffe, E. H., Li, C., Thompson, E. A., Maller, J. L., and Sluder, G. (1999)
Science 283, 851-854
8. Lacey, K. R., Jackson, P. K., and Stearns, T. (1999) Proc. Natl. Acad. Sci. USA 96,
2817-2822
9. Koff, A., Giordano, A., Desai, D., Yamashita, K., Harper, J. W., Elledge, S.,
Nishimoto, T., Morgan, D., Franza, B. R., and Roberts, J. M. (1992) Science 257,
1689-1694
10. Dulic, V., Lees, E., and Reed, S. (1992) Science 257, 1958-1961
11. van der Heuvel, S., and Harlow, E. (1993) Science 262, 2050-2054
12. Ohtsubo, M., and Roberts, J. M. (1993) Science 259, 1908-1912
13. Mussman, J. G., Horn, H. F., Carroll, P. E., Okuda, M., Tarapore, P., Donehower, L.
A., and Fukasawa, K. (2000) Oncogene 19, 1635-46
14. Resnitzky, D., Gossen, M., Bujard, H., and Reed, S. I. (1994) Mol. Cell. Biol. 14,
52 1669-1679
15. Ohtsubo, M., Theodoras, A. M., Schumacher, J., Roberts, J. M., and Pagano, M.
(1995) Mol. Cell. Biol. 15, 2612-24
16. Okuda, M., Horn, H. F., Tarapore, P., Tokuyama, Y., Knudsen, E. S., Hofmann, I. A.,
Snyder, J. D., Bove, K. E., and Fukasawa, K. (2000) Cell. 103, 127-140
17. Prestayko, A. W., Klomp, G. R., Schmoll, D. J., and Busch, H. (1974) Biochemistry
13, 1945-51
18. Spector, D. L., Ochs, R. L., and Busch, H. (1984) Chromosoma 90, 139-48
19. Yung, B. Y., Busch, H., and Chan, P. K. (1985) Biochim Biophys Acta 826, 167-73
20. Borer, R. A., Lehner, C. F., Eppenberger, H. M., and Nigg, E. A. (1989) Cell 56, 379-
90
21. Takemura, M., Sato, K., Nishio, M., Akiyama, T., Umekawa, H., and Yoshida, S.
(1999) J. Biochem (Tokyo) 125, 904-9
22. Szebeni, A., and Olson, M. O. (1999) Protein Sci. 8, 905-12
23. Higuchi, R. PCR Technology (Erlich, H. A. ed.), Stockton Press, New York, pp. 61-
70
24. Fukasawa, K., Wiener, F., Vande Woude, G. F., and Mai, S. (1997) Oncogene 15,
1295-1302
25. Desai, D., Wessling, H. C., Fisher, R. P., and Morgan, D. O. (1995) Mol. Cell. Biol.
15, 345-350
26. Fukasawa, K., Murakami, M. S., Blair, D. G., Kuriyama, R., Hunt, T., Fischinger, P.,
and Vande Woude, G. F. (1994) Cell Growth Differ. 5, 1093-1103
27. Pines, J., and Hunter, T. (1990) Nature 346, 760-763
53 28. Marraccino, R. L., Firpo, E. J., and Roberts, J. M. (1992) Mol. Biol. Cell. 3, 389-401.
29. Carbonaro-Hall, D., Williams, R., Wu, L., Warburton, D., Zeichner-David, M.,
MacDougall, M., Tolo, V., and Hall, F. (1993) Oncogene 8, 1649-1659
30. Pagano, M., Pepperkok, R., Verde, F., Ansorge, W., and Draetta, G. (1992) EMBO J.
11, 961-971
31. Nurse, P. (1994) Cell 79, 547-550
32. Peter, M., Nakagawa, J., Doree, M., Labbe, J. C., and Nigg, E. A. (1990) Cell 60,
791-801
33. Jiang, P. S., Chang, J. H., and Yung, B. Y. M. (2000) Cancer Lett. 153, 151-160
34. Chan, P. –K., Liu, Q. R., and Durban, E. (1990) Biochem. J. 270, 549-552
35. Joshi, H. C. (1994) Curr. Opin. Cell Biol. 6, 54-62
36. Pihan, G. A., Purohit, A., Wallace, J., Knecht, H., Woda, B., Quesenberry, P., and
Doxsey, S. J. (1998) Cancer Res. 58, 3974-3985
37. Lingle, W. L., and Salisbury, J. L. (1999) Am. J. Pathol. 155, 1941-1951
38. Weber, R. G., Bridger, J. M., Benner, A., Weisenberger, D., Ehemann, V.,
Reifenberger, G., and Lichter, P. (1998) Cytogen. & Cell Gen. 83, 266-269
39. Carroll, P. E., Okuda, M., Horn, H. F., Biddinger, P., Stambrook, P.J., Gleich, L. L.,
Li, Y.-Q., Tarapore, P., and Fukasawa, K. (1999) Oncogene 18, 1935-1944
40. Dyson, N. (1998) Genes & Dev. 12, 2245-2262
41. Nevins, J. R. (1998) Cell Growth Differ. 9, 585-593
42. Meraldi, P., Lukas, J., Fry, A. M., Bartek, J., and Nigg, E. A. (1999) Nature Cell Biol.
1, 88-93
43. Balczon, R, Bao, L, Zimmer WE, Brown K, Zinkowski RP and Brinkley BR. (1995)
54 J. Cell Biol. 130, 105-115
44. Zatsepina, O. V., Rousselet, A., Chan, P. K., Olson, M. O., Jordan, E. G., and
Bornens, M. (1999) J. Cell Sci. 112, 455-466
45. Lewin, B. (1990) Cell 61, 743-752
46. Nigg, E. A. (1988) Int. Rev. Cytol. 110, 27-92
47. Rappaport, R. (1986) Int. Rev. Cytol. 105, 245-281
55 Figure Legends
Figure 1. CDK2/cyclin E phosphorylates NPM/B23 in vitro on threonine residue(s).
A. Wild-type NPM/B23 and NPM mutant fused to GST [GST-NPM/wt and GST-
NPM(∆186-239), respectively] as well as GST were subjected to in vitro kinase reactions with CDK2-cyclin E. The reaction samples were run on 10% SDS-PAGE (pH 8.8), and autoradiographed (left panel). The right panel shows the Coomassie blue-stained gel.
Lane 1, GST; lane 2, GST-NPM∆186-239; lane 3, GST-NPM/wt).
B. The 32P-labeled GST-NPM/wt proteins as shown in (A) were purified from the gel, and subjected to acid hydrolysis. The phosphorylated amino acids were then separated by 2D- electrophoresis and visualized by autoradiography as described previously (27).
The position of the migrations of phosphoserine, phosphothreonine, and phosphotyrosine standards, detected by ninhydrin staining are indicated by circles. The sample loading origin is indicated by +.
Figure 2. NPM/B23 is phosphorylated by CDK2/cyclin E on threonine 199 (Thr199) in vitro.
A. Amino acid sequence of human NPM/B23 between the residue 186 and 239. All threonine residues in this region were indicated by arrowheads. Each threonine residues are followed by proline, which is known to comprise a CDK phosphorylation consensus sequence (45, 46). By PCR-assisted mutagenesis, each threonine residue was replaced by alanine as indicated.
B. GST-NPM/wt (wt) and four point mutation NPM/B23 fusion proteins (T199A, T219A,
T234A, and T237A) were subjected to in vitro kinase reactions with either CDK2/cyclin
E (top panels) or CDK2/cyclin A (bottom panels). As a negative control, GST was used
56 as a substrate (lane 1). Among these mutants, GST-NPM/T199A failed to be
phosphorylated by both CDK2/cyclin E and CDK2/cyclin A (lane 3). In contrast, other
mutant GST-NPM fusion proteins (lane 4-6) were phosphorylated by both CDK2/cyclin
E and CDK2/cyclin A at similar efficiency with wild-type NPM (lane 2). Coomassie
blue-staining of the gel is shown on the right for each set.
C. CDK1/cyclin B phosphorylates NPM/B23 on Thr234 and Thr237. Wt and mutant
GST-NPM as well as GST-NPM double alanine substitution mutant (Thr234→A and
Thr237→A; T234A/T237A) were subjected to in vitro kinase reactions with
immunopurified CDK1/cyclin B from mitotic Swiss 3T3 cells using anti-cyclin B
antibody (GNS1, Santa Cruz). The level of 32P-incorporation was quantitated by
scanning with Fuji Phosphoimager 1000, and the result is presented in arbitrary units with
the level of 32P-incorporation observed in GST-NPM/wt as 1.0. The amounts of GST-
NPM proteins used in the reactions (2 µg/30 µl) were confirmed to be approximately equal by Coomassie blue-staining of the gel (not shown).
Figure 3. The site of NPM/B23 phosphorylated in vitro by CDK2/cyclin E is
phosphorylated in vivo.
A. Immunoblot analysis of cyclin E expression in HeLa cells stably transfected with
human cyclin E. HeLa cells were co-transfected with a plasmid encoding human cyclin E
(or an empty vector as a control) and a plasmid encoding a neomycin resistant gene. The
G418-resistant colonies from the cyclin E and the vector transfection were sub-cloned
(HeLa/CycE and HeLa/Vec, respectively). The extracts prepared from these cells (50 µg
of total protein) were subject to immunoblot analysis using anti-human cyclin E
57 polyclonal antibody (sc-198, Santa Cruz). HeLa/CycE cells expresses ∼5-fold more
cyclin E proteins than HeLa/Vec cells.
B. Histone H1 kinase activity of HeLa/CycE cells. The cell extracts derived from
exponentially growing HeLa/CycE and the control HeLa/Vec cells were
immunoprecipitated with anti-human cyclin E antibody, and the immunoprecipitates were
subjected to an in vitro histone H1 kinase assay as described previously (13).
C. Tryptic phosphopeptide mapping of in vitro phosphorylated NPM/B23 by
CDK2/cyclin E and in vivo phosphorylated NPM/B23 in HeLa/CycE cells. In vitro phosphorylated GST-NPM/wt by CDK2/cyclin E was prepared as described in the legend to Fig. 1. For preparation of in vivo phosphorylated NPM/B23, HeLa/CycE cells were metabolically labeled in the presence of 32P-orthophosphate. The lysates were
immunoprecipitated with anti-NPM/B23 monoclonal antibody, and the
immunoprecipitates were resolved in 10% SDS-PAGE. 32P-labeled NPM/B23 proteins
were eluted from the gel. The in vitro 32P-labeled GST-NPM/wt (panel a) and in vivo
32P-labeled NPM/B23 (panel b) were oxidized with performic acid, and digested with
trypsin as described previously (27). Each tryptic digestion sample was loaded onto a
thin layer cellulose gel plate, and subjected to electrophoresis (horizontal dimension),
followed by ascending chromatography. Mixed map was generated by loading equal
counts of trypsin-digested in vitro 32P-labeled GST-NPM/wt and in vivo 32P-labeled
NPM/B23 (panel c). The origin of the sample placement is indicated by +. The arrows
indicate the 32P-labeled tryptic fragment which is observed both in the in vitro and in vivo
samples.
58 Figure 4. Initiation of centrosome duplication is blocked by NPM/T199A expression.
A. Immunoblot analysis of cells transfected with wild-type NPM/B23 (NPM/wt) and
NPM/T199A. Swiss 3T3 cells were transiently transfected with plasmids encoding
FLAG epitope-tagged NPM/wt and NPM/T199A. As a control, an expression vector used for construction of the plasmids was transfected. For each transfection, a plasmid encoding a puromycin resistant gene (pBabe/puro) was co-transfected as a selection marker. Puromycin was added to culture medium 16 h after transfection. Cells that had been successfully transfected were selected within 36 h after addition of puromycin. The puromycin- selected cells were replated, and further cultured for additional 24 h. The whole cell lysates were prepared from the transfectants, and probed with anti-FLAG polyclonal antibody. NPM/wt and NPM/T199A-transfectants expressed similar levels of transfected NPM/B23.
B & C. The transfectants described above in (A) were immunostained for centrosomes
(centrioles). After cold-treatment and brief extraction (see Experimental Procedures), cells were fixed and co-immunostained with anti-γ-tubulin polyclonal and anti-α-tubulin monoclonal (DM1A) antibodies. Antigen-antibody complexes were detected with FITC- conjugated goat anti-rabbit IgG and rhodamine-conjugated goat anti-mouse IgG antibodies. The nuclei were also counterstained with DAPI. The representative immunostainings are shown in (C). Panels a & e, γ-tubulin immunostaining; panels b & f, α-tubulin immunostaining; panels c & g, DAPI staining; panels d & h, overlay images of γ-tubulin and α-tubulin immunostainings. Panels I-VIII ∼x5 magnification of the corresponding centrosome images indicated in panels b & f. At a higher magnification, each single dot detected by anti-γ-tubulin antibody was resolved to doublets (representing
59 a pair of centrioles) by anti-α-tubulin antibody. Anti-γ-tubulin antibody reactive doublets
(potentially duplicated centrosomes) are indicated by arrowheads, and anti-γ-tubulin
antibody reactive singlets (unduplicated centrosomes) are indicated by arrows. Scale bars
for the images shown in panels a-h, 20 µm. The number of centrosomes per cells (n)
were categorized into n=1, n=2, and n≥3. For each transfectant, >400 cells were
examined. The results shown in (B) are the average centrosome profiles determined from
three independent experiments.
Figure 5. Expression of NPM/T199A mutant specifically inhibits centrosome
duplication.
Swiss 3T3 cells were transiently transfected with either a plasmid encoding FLAG epitope-tagged NPM/T199A or a control vector. For each transfection, pBabe/puro was co-transfected as a selection marker. Puromycin was added to medium 16 h after transfection. Puromycin-resistant cells at 36 h after addition of puromycin were replated, and further cultured for 24 h. During the final 3 h of culturing, BrdU was added to medium. Cells were then processed for co-immunostaining with anti-BrdU monoclonal
(B: panels a & b) and anti-γ-tubulin polyclonal antibodies (B: panels a’ & b’). First,
percent of cells which had incorporated BrdU were determined through examination of
>300 cells (A). In vector-transfected cells, virtually all of the BrdU-positive cells
contained two anti-γ-tubulin antibody-reactive dots (duplicated centrosomes) (B; panels a
& a’, indicated by arrows), while majority of BrdU-positive NPM/T199A contained a
single dot (unduplicated centrosome) (B; panels b & b’, indicated by an arrow). The
number of centrosomes per cell in the BrdU-positive cells was scored by fluorescence
60 microscopy. For each transfectant, >100 BrdU-positive cells were examined, and the results from three independent experiments are shown in (C).
Figure 6. Expression of NPM/T199A mutant results in a high frequency of monopolar mitosis.
Swiss 3T3 cells were transiently co-transfected with either a plasmid encoding FLAG epitope-tagged NPM/T199A or a control vector along with pBabe/puro as a selection marker. Puromycin-resistant cells were selected as described in the legend to Fig. 4.
Cells were immunostained with anti-α- and β- tubulin antibodies (B; panels b and f) and anti-γ-tubulin polyclonal antibody (B; panels a and e). Cells were also counterstained with DAPI (B; panels c and g). Antigene-antibody complexes were visualized by FITC- conjugated goat anti-mouse IgG and rhodamine-conjugated goat anti-rabbit IgG antibodies. The mitotic cells were first identified by condensed chromosomes under a fluorescence microscope, and the number of spindle poles in each mitotic cells were determined (A). For each transfectant, >50 mitotic cells were examined. Representative immunostaining images are shown in (B). The panels d and h are overlay images of panels a-c and e-g, respectively. The arrows point to the spindle pole. Scale bar, 10 µm.
Figure 7. NPM/T199A physically associates with centrosomes.
The control vector-transfected (panels A-D) and FLAG-epitope tagged NPM/T199A- transfected cells (panels E-H) described in the legend to Fig. 4 were examined for sub- localization of transfected NPM/B23 mutant proteins by co-immunostaining with anti-γ- tubulin (green, A & E) and anti-FLAG monoclonal (red, B & F) antibodies. Cells were also counterstained with DAPI (C & G). Panels D and H show the overlay images. The
61 arrow points to the centrosome. The arrow in panel B was placed at the same position as shown in panel A. Scale bar, 10 µm.
62 A 123 123
GST-NPM/wt GST-NPM (∆186-239)
GST
In vitro kinase assay Coomassie blue stain
B
pi
p-Ser p-Thr
p-Tyr
+
Figure 1
63 A
186 239 AEEKAPVKKSIRDTPAKNAQKSNQNGKDSKPSSTPRSKGQESFKKQEK TPKTPK
AAAA (T199A) (T219A) (T234A) (T237A)
B 219A 199A 234A 237A 199A 219A 234A 237A T GST Wt T T T kDa GST Wt T T T T 75 GST-NPM 50 CDK2/cyclin E 35
25 GST 123456 123456 199A 219A 234A 237A 199A 219A 234A 237A GST Wt T T T T GST Wt T T T T 75 GST-NPM 50 CDK2/cyclin A 35
25 GST 123456 123456 In vitro kinase assay Coomassie blue staining
C 7A 23 /T A A A A A 9 9 4 7 34 CDK1/cyclin B 9 1 23 3 2 t 1 2 2 T W T T T T GST-NPM
1
ts 0.8
y uni 0.6 ar
bitr 0.4 Ar 0.2
0 123456
Figure 2
64 AB /Vec /Vec /CycE /CycE HeLa HeLa HeLa HeLa kDa 118
90 cyclin E 70 Histone H1 55
38
C a) in vitro b) in vivo c) in vitro + in vivo
4 5 4 5
2 2 romatography 1 3 h 3 1 C ++ + − + Electrophoresis
Figure 3
65 A B 100 one centrosome 9A two centrosomes t 19 80 T > two centrosomes r /w / o M M ct P P e N N V lls
e 60
FLAG- 40 NPM/B23
Percent of c 20
0 C Vector NPM/wt NPM/T199A abI II
III II I
IV
cdIII
IV
V efV
VI
VI
VII VIII
ghVII
VIII
Figure 4
66 Figure 5
67 Figure 6
68 A B
Anti-γ-tubulin Anti-FLAG
C D
DAPI Overlay
E F
Anti-γ-tubulin Anti-FLAG
G H
DAPI Overlay
Figure 7
69
Chapter III
CDK2/CyclinE-Mediated Phosphorylation on Threonine 199 of Nucleophosmin/B23 Localizes to Nuclear Speckles and Represses Pre-mRNA Splicing
Yukari Tokuyama, Akila Mayeda, Pheruza Tarapore and Kenji Fukasawa
Department of Cell Biology, University of Cincinnati College of Medicine PO Box 670521, Cincinnati, OH 45267-0521
70 Abstract
Nucleophosmin (NPM/B23) is a multi-functional phosphoprotein predominantly
localized in the nucleolus, and is thought to function in ribosome assembly,
cytoplasmic/nuclear shuttling, control of centrosome duplication and molecular
chaperoning. NPM/B23 is phosphorylated by several kinases, including nuclear kinase
II, casein kinase II, and cyclin-dependent kinases (CDK2 and CDK1). These
phosphorylations appear to determine the function of NPM/B23. Previously, we
identified Thr199 of NPM/B23 as a major phosphorylation site mediated by CDK2, and found that this phosphorylation was involved in regulation of centrosome duplication. In this study, we further examined the effect of CDK2-mediated phosphorylation of
NPM/B23 by using an antibody that recognizes NPM/B23 that is phosphorylated on
Thr199. We found that the phospho-Thr199 NPM specifically localized to dynamic
nuclear structures known as nuclear speckles, which consist of pre-mRNA splicing
factors. Also, we found that phosphorylation on Thr199 by CDK2/cyclin E enhances the
RNA-binding activity of NPM/B23. Moreover, phospho-Thr199 NPM repressed pre-
mRNA splicing activity. These findings suggest the role of CDK2-mediated
phosphorylation on Thr199 of NPM/B23 in cell cycle regulated control of pre-mRNA
splicing.
71 Introduction
Nucleophosmin (NPM), also known as B23, NO38 or numatrin, is a phosphoprotein abundantly found in the nucleolus. NPM/B23 is a multifunctional protein, and has been implicated in a variety of cellular events. Ultrastructural studies have shown that NPM/B23 specifically localizes at the granular regions of the nucleoli
(1). In addition, NPM/B23 binds RNA, and possesses an endoribonuclease activity (2-4).
These observations have suggested that NPM/B23 is involved in processing of pre- ribosomal RNA. NPM/B23 has also been shown to control DNA duplication through forming complexes with DNA polymerase α and RB (5). It has also been shown that
NPM/B23 plays an important role in protein trafficking between the cytoplasm and nucleus through directly binding to the nuclear localization signals (NLS) of the target proteins (6-10). NPM/B23 is also found at centrosomes, and the centrosomal association of NPM/B23 is thought to be an important factor controlling centrosome duplication (11).
Moreover, NPM/B23 has been shown to possess molecular chaperoning activity. It can prevent heat-induced protein aggregation, and can promote renaturation of chemically as well as heat denatured proteins (12).
NPM/B23 is phosphorylated by several different kinases, including casein kinase
II (CKII), nuclear kinase II, polo-like kinase 1 and cyclin-dependent kinases
(CDK1/cyclin B, CDK2/cyclin E, and CDK2/cyclin A) (13-20). Phosphorylation of
NPM/B23 by CKII has been shown to increase its affinity for the NLS sequences present on the SV40 large T antigen as well as HIV Rev protein (9). In addition, it has also been shown that phosphorylation of NPM/B23 by CKII releases denatured proteins from complex of NPM/B23-denatured protein substrate (16). We and others have recently
72 shown that NPM/B23 is phosphorylated on threonine 199(Thr199) by CDK2/cyclin E
(19,20), a kinase complex activated specifically in late G1 phase (21,22), and this
phosphorylation affects the affinity of NPM/B23 to centrosomes (11).
Nuclear speckles primarily consist of pre-mRNA splicing factors, and correspond
to electron microscopically identifiable structures called the perichromatin fibrils (PFs)
that stem out from and surround the interchromatin granule clusters (IGCs) (23).
Although the exact function of these nuclear speckles is not fully understood, previous
studies suggest their involvement in the storage and/or assembly of pre-mRNA splicing
factors (24).
In this communication, we further exploited the effect of the Thr199
phosphorylation on the activity of NPM/B23 by the use of an antibody, which
specifically recognizes NPM/B23 phosphorylated on Thr199 (phospho-Thr199 NPM). We found that phospho-Thr199 NPM specifically localizes at intranuclear structures known as
“nuclear speckles” (23). Also, as expected from the cell cycle phase-specific activation
of CDK2/cyclin E, the speckled distribution of phospho-Thr199 NPM occurs in a cell
cycle-dependent manner. Upon inhibition of transcription, phospho-Thr199 NPM (along
with other splicing factors) becomes sequestered to enlarged and rounded speckles.
Moreover, we also find that the phosphorylation on Thr199 enhances the RNA-binding
activity of NPM/B23. Furthermore, the CDK2/cyclin E-mediated phosphorylation on
Thr199 of NPM/B23 repressed pre-mRNA splicing activity. These findings strongly
suggest the role of the CDK2/cyclin E-mediated phosphorylation of NPM/B23 in
pre-mRNA processing.
73 Materials and Methods
Cells, Transfection and Drug Treatment.
Wild-type mouse skin fibroblasts (MSFs) were prepared from abdominal skins of an 8-
week-old C57L male mouse. MSFs were maintained in complete medium [Dulbecco
Modified Eagle Medium supplemented with 10% fetal bovine serum (FBS), penicillin
(100 U/ml) and streptomycin (100 µg/ml)], in an atmosphere containing 10% CO2.
Transient transfection of RNPS1 sequence was performed as described previously (19).
Briefly, MSFs were co-transfected with plasmid encoding T7-tagged RNPS1 (25) with a puromycin resistance gene plasmid (pBabe/puro) at a molar ratio of 20:1 by calcium phosphate protocol. The puromycin-resistant cells were pooled and replated on cover slips and further cultured in fresh complete medium for 24 h. For inhibition of transcription, α-amanitin (50 µg/ml) (Sigma-Aldrich) was added to the cells for 5 h.
Purification of GST-NPM/B23.
Wild-type and mutant NPM/B23 cDNA sequences were fused to glutathione S-
transferase using a two-step PCR as described (26). The PCR products were inserted in
frame into a pGEX-4T-1 vector using BamH1 and EcoR1 restriction sites. GST-NPM
fusion proteins were bacterially purified according to the protocol provided by the
manufacturer (Amersham Pharmacia Biotech). Briefly, cells were induced with isopropyl-1-thio-β-D-galactopyranoside for 4 h. Cells were harvested and clarified
bacterial lysates were passed through a Glutathione Sepharose 4B column, and GST-
NPM proteins were eluted. The concentration of the eluted GST-NPM was estimated by
comparison with a known concentration of bovine serum albumin run in parallel on SDS-
PAGE.
74 Immunoblot Analysis.
Cells were washed three times with phosphate-buffered saline (PBS) and lysed in
SDS/Nonidet P-40 lysis buffer [1% SDS, 1% Nonidet P-40, 50 mM Tris (pH 8.0), 150
mM NaCl, 4 mM Pefabloc SC (Roche Molecular Biochemicals), 2 µg/ml leupeptin, 2
µg/ml aprotinin]. The lysates were boiled for 5 min, and then cleared by a 10 min centrifugation at 20,000 x g at 4oC. The supernatant was denatured at 95oC for 5 min in
sample buffer [2% SDS, 10% glycerol, 60 mM Tris (pH 6.8), 5% -mercaptoethanol,
0.01% bromophenol blue]. Samples were resolved by SDS-PAGE, and transferred onto
Immobilon-P sheets (Millipore). The blots were incubated in blocking buffer [5%
(wt/vol) nonfat dry milk in Tris-buffered saline (TBS) + Tween 20 (TBS-T)] for 1 h at
room temperature. The blots were then incubated with primary antibody for overnight at
4oC, followed by incubation with horseradish peroxidase-conjugated secondary antibody for 1 h at room temperature. The antibody-antigen complex was visualized by ECL chemiluminescence (Amersham Pharmacia Biotech).
Indirect Immunofluorescence.
Cells grown on coverslips were fixed with 10% formalin/10% methanol for 20 min at
room temperature. The cells were permeabilized with 1% Nonidet P-40 in PBS for 5
min, followed by incubation with blocking solution [10% normal goat serum in PBS] for
1 h. Cells were then probed with primary antibodies for 1 h, and antibody-antigen
complexes were detected with either Alexa Fluor 488- or Alexa Fluor 594-conjugated
goat secondary antibody (Molecular Probes) by incubation for 1 h at room temperature.
The coverslips were washed three times with PBS after each incubation, and then
counterstained with 4',6-diamidino-2-phenylindole (DAPI) for 10 min at room
75 temperature. Immunostained cells were examined under a fluorescence microscope
(Zeiss Axioplan 2 Imaging, 60X objective lens) or confocal microscope (Zeiss LSM510,
63X objective lens).
Alkaline Phosphatase Treatment.
Cells grown on coverslips were fixed with 100% methanol for 10 min at -20°C. Cells were then air-dried, and rehydrated in PBS for 10 min at room temperature. Cells were then incubated in the solution [100 mM glycine (pH10.4)] containing 10 units of alkaline phosphatase type IV (Sigma-Aldrich) for 2 h at 37°C. The control cells were incubated in the solution without alkaline phosphatase.
In Vitro Kinase Assay.
For in vitro phosphorylation of NPM/B23 by CDK2/cyclin E or CKII, GST, GST-
NPM/wt, or GST-NPM/T199A were incubated with baculovirally purified active
CDK2/cyclin E (27) or CKII (New England Biolabs). The in vitro kinase reactions were
performed in 10 mM PIPES buffer in the presence of ATP at 25°C for 15 min, and at
32°C for additional 15 min. The samples were resolved by SDS-PAGE, and followed by immunoblot analysis as described above.
BrdU Incorporation Assay.
The assay was performed using the BrdU-labeling kit (Roche Molecular Biochemicals)
according to the manufacturer’s instruction. Briefly, cells were fixed in 70% ethanol in 50
mM glycine (pH 2.0) for 20 min at -20°C and then probed with anti-BrdU monoclonal
antibodies for 45 min at 37°C. Antigen-antibody complexes were detected by FITC-
conjugated sheep anti-mouse IgG.
76 RNA Binding Assay.
The assay was performed as described by Okuwaki et al. (20,28). Briefly, RNA was extracted from MSFs using TRIZOL (GibcoBRL). GST-NPM/B23 proteins were subjected to an in vitro kinase assay with CDK2/cyclin E as described above, and then mixed with RNA for 30 min at room temperature. The samples were loaded onto 15-
40% sucrose gradient [20 mM Tris (pH7.4), 50 mM NaCl, 0.5 mM PMSF, and 1 mM dithiothreitol] and centrifuged at 39,000 rpm for 4 h and fractions were collected from the bottom. The fractions were run on 10% SDS-PAGE for immunoblot analysis with anti-
NPM/B23 antibody, and on 1% agarose-formaldehyde gel for northern blot analysis using 32P-labelled 18S rRNA DNA probe (Ambion).
Immunoprecipitation.
Cells were washed with PBS three times, and then lysed in lysis buffer [150mM NaCl, 50 mM HEPES (pH7.0), 0.1%NP-40 and a cocktail of inhibitors] as described in (29). The lysates were cleared by centrifugation for 15 min at 20,000 g at 4 °C. The lysates containing 400 µg of total proteins were immunoprecipitated with antibodies described in text. The antibody-antigen complexes were collected with either protein G- or A- agarose, and washed three times with lysis buffer. The immunocomplexes were then denatured in sample buffer at 95 °C for 5 min and resolved by 10% SDS-PAGE.
In Vitro Splicing Assay. m7GpppG-capped 32P-labeled pre-mRNA fragment was made by runoff transcription of linearlized β-globin template DNA with SP6 RNA polymerase (30,31). This transcript was used as a substrate for splicing assay. Indicated amounts of mock or CDK2/cyclin E phosphorylated GST, GST-NPM/wt, or GST-NPM/T199A was added to splicing
77 reactions (25 µl) with HeLa nuclear extract or S100 extract with recombinant SF2/ASF
(30,32)), and 20 fmol of 32P-labeled β-globin pre-mRNA and incubated at 30 C° for 2-3 h
(33). The spliced products were analyzed by 5.5% polyacrylamide/7M urea gel and autoradiography.
78 Results
Specificity of Phospho-Thr199 NPM Antibody. It has previously been shown that
NPM/B23 is phosphorylated primarily on Thr199 by CDK2/cyclin E (19), and this
phosphorylation is important in the regulation of centrosome duplication (11). To further
analyze the significance of the Thr199 phosphorylation in other cellular events, we
obtained an antibody, which specifically reacts with NPM/B23 phosphorylated on
Thr199. We first tested the specificity of the anti-phospho-Thr199 NPM antibody by examining its reactivity to phosphorylated NPM/B23. The bacterially purified wild-type
NPM/B23 proteins fused to GST (GST-NPM/wt) were subjected to an in vitro kinase
assay with CDK2/cyclin E. Since casein kinase II (CKII) has been shown to be another
major kinase that phosphorylates NPM/B23 on the residues other than Thr199 (14,16), in
vitro kinase reaction was also performed with CKII as a control. We also included a
NPM/B23 mutant whose Thr199 residue was replaced with alanine (GST-NPM/T199A)
as well as GST alone as controls. The reaction samples were resolved by SDS-PAGE,
and immunoblotted using phopsho-Thr199 NPM antibody (Fig 1A, left panel). To
demonstrate that all of the reaction samples contain similar levels of either GST-NPM/wt
or GST-NPM/T199A, the same blot was stained with Commassie Blue (Fig 1A, right
panel). GST-NPM/wt phosphorylated by CDK2/cyclin E was readily detected by the
anti-phospho-Thr199 NPM antibody (lane 8). It should be noted that the GST-NPM/wt
phosphorylated by CDK2/cyclin E showed a slightly faster migration than the
unphosphorylated form (lane 17), which is consistent with the previous studies (11). In
contrast to wild-type GST-NPM, neither GST (lane 7) nor GST-NPM/T199A (lane 9)
was recognized by the anti-phospho-Thr199 NPM antibody. Phosphorylation by CKII on
79 GST-NPM/wt (lane 5) and GST-NPM/T199A (lane 6) was also not detected by this
antibody. Thus, the anti-phospho-Thr199 NPM antibody is specific to NPM/B23
phosphorylated on Thr199 by CDK2/cyclin E.
Phospho-Thr199 NPM localizes in speckled pattern in the nuclei. We next examined
whether phospho-Thr199 NPM shows any specific subcellular localization by
immunostaining primary skin fibroblasts (MSFs) derived from adult mice with the anti-
phospho-Thr199 NPM antibody. As a control, we immunostained MSFs with anti-
NPM/B23 antibody, which is raised against purified NPM/B23 proteins and reacts with
both phosphorylated and non-phosphorylated forms of NPM/B23 (34). The anti-
NPM/B23 antibody detected NPM/B23 primarily in the nucleolus and to lesser extent in
the nucleoplasm as scattered small dot-like structures (Fig. 1B, panels a-c). In contrast,
anti-phospho-Thr199 NPM antibody did not detect NPM/B23 in the nucleolus, but
detected it in distinct intranuclear structures as well as in diffuse regions throughout the
nucleoplasm (Fig 1B, panels d-f). The localization in a speckled pattern detected by the
anti-phospho-Thr199 NPM antibody was clearly distinct from nucleoli, visualized using
the differential interference contrast microscopy (Fig. 1B, panel e).
To confirm the specificity of the immunostaining pattern of the anti-phospho-
Thr199 NPM antibody to the phosphorylated form of NPM/B23, we treated the fixed
MSFs with alkaline phosphatase and immunostained with the anti-phospho-Thr199 NPM antibody. The control untreated MSFs (Fig. 1C, panel a) showed the presence of phospho-Thr199 NPM in IGCs and PFs as shown in Fig. 1B (panel d). However, the anti-
phospho-Thr199 NPM antibody failed to detect antigens in the alkaline phosphatase
80 treated cells, demonstrating that this antibody is indeed detecting the phosphorylated
epitope (Fig. 1C, panel c).
The levels of phospho-Thr199 NPM during the cell cycle. Since CDK2/cyclin E is
activated specifically in mid-late G1 phase of the cell cycle in mouse cells, it is likely that
cellular levels of phospho-Thr199 NPM change during the cell cycle. To test this
possibility, we synchronized MSFs by serum starvation, followed by serum stimulation.
At every 5 h after serum-stimulation for a period of 20 h, cells were fixed and examined
for localization of phospho-Thr199 NPM by immunostaining (Fig. 2A). To monitor the
cell cycle progression, cells were cultured in parallel in the presence of BrdU (Fig. 2C).
Under a serum-starved condition, there was little or no phospho-Thr199 NPM staining as
expected from the fact that no active CDK2/cyclin E is available. Between 5 and 10 h
after serum stimulation, weak yet readily detectable signal for phospho-Thr199 NPM
staining were observed in cells. At 20 h, signal intensity of phospho-Thr199 NPM staining
increased in the majority of cells. To quantitate the cell cycle-dependent phospho-Thr199
NPM signal, the images taken by confocal microscopy were subjected to computational
analysis by morphometric measurement using Metamorph Software (Universal Imaging
Corp). The intensity of the phospho-Thr199 NPM signal was calculated (Fig. 2B). We
found that the intensity of the phospho-Thr199 NPM signal gradually increased during
serum stimulation, and importantly the rate of the increase strongly paralleled the rate of
BrdU incorporation (Fig. 2C). Considering that CDK2/cyclin E is activated prior to S
phase entry, these results suggest that appearance of phospho-Thr199 NPM in the nuclear
speckle pattern is primarily mediated by activated CDK2/cyclin E.
81 To verify the immunocytochemical observations, we examined the phospho-
Thr199 NPM protein level in MSFs after serum stimulation for 0, 5, 10, 15, 20 h by
immunoblot analysis (Fig. 2D, left panel). At 0 h, no phospho-Thr199 NPM was detected.
As the cell cycle progressed (5, 10, 15 h), we detected low levels of phospho-Thr199
NPM. At 20 h of serum stimulation, the level of phospho-Thr199 NPM increased dramatically. These results are consistent with the changes in the level of phospho-Thr199
NPM during G1 progression observed immunocytochemically. To exclude the possibility that the changes in the level of phospho-Thr199 NPM are due to the changes in the NPM/B23 expression level, the same blot was re-probed with anti-NPM/B23 antibody, which recognizes both non-phosphorylated and phosphorylated NPM/B23 (Fig.
2D, center panel). We found that there was no significant change in the total NPM/B23 protein levels during 20 h of serum stimulation.
Phospho-Thr199 NPM localizes to nuclear speckles. The observation of the unique
subnuclear localization of phospho-Thr199 NPM prompted us to search other nuclear
compartments with similar localization patterns. The pre-mRNA splicing machinery,
which consists of small nuclear ribonucleoprotein particles (snRNPs) along with non- snRNP protein factors have a similar intranuclear speckle pattern known as nuclear
speckles. Splicing factor SC-35, a non-snRNP spliceosome component that is essential
for spliceosome assembly and spliceosome function, has been used as a hallmark protein
for nuclear speckles (35,36). Thus, we co-immunostained using anti-phospho-Thr199
NPM polyclonal antibody and anti-SC-35 monoclonal antibody (Fig. 3A). We found that the phospho-Thr199 NPM and the SC-35 staining signals showed a high degree of co-
82 localization in certain speckled regions. This staining pattern is also characteristic for
many other splicing factors (37).
Previous studies involving the biochemical purification of an activity that
stimulates pre-mRNA splicing from HeLa nuclear extracts co-purified NPM/B23 along
with a general splicing activator, RNPS1 (25). To determine whether RNPS1 co-
localizes with phospho-Thr199 NPM, we transiently expressed T7-tagged RNPS1 in MSFs
and examined cells for phospho-Thr199 NPM (anti-phospho-Thr199 NPM polyclonal
antibody) and RNPS1 (anti-T7 monoclonal antibody (Novagen)) (Fig. 3B). Phospho-
Thr199 NPM and RNPS1 co-localized in an intense speckled pattern, very similar to the
SC-35 localization pattern. These experiments suggest the potential role of phospho-
Thr199 NPM in the pre-mRNA processing.
Nuclear speckles are structurally dynamic, with the constituent proteins changing
depending on the transcriptional activity of the cell. Upon transcription activation, pre-
mRNA splicing factors such as SC-35 are recruited from IGCs to PFs (23,38). It has
been shown that inhibition of transcription results in a subnuclear redistribution of
proteins involved in pre-mRNA processing and splicing (23,39). Administration of α-
amanitin, a commonly used transcription inhibitor, which specifically inhibits RNA
polymerase II (40), results in appearance of enlarged rounded speckles of splicing factors
and disappearance of diffuse connections between the speckles in the nucleoplasm (41).
To explore the potential significance of phospho-Thr199 NPM in pre-mRNA processing,
we tested whether phospho-Thr199 NPM would redistribute to enlarge speckles upon transcription inhibition in a similar manner as is observed with splicing factors such as
SC-35. Exponentially growing MSFs were treated with α-amanitin for 5 h, and then co-
83 immunostained with anti-phospho-Thr199 NPM and anti-SC-35 antibodies. In the control
untreated cells, we observed the characteristic IGCs and PFs stained speckle pattern of
SC-35 and phospho-Thr199 NPM (Fig. 3C, panels a-d). In the α-amanitin treated cells,
SC-35 redistributed to enlarged rounded speckles (IGCs) and the diffused staining (PFs)
was no longer detected (panel f) (41). Moreover, in these cells the phospho-Thr199 NPM co-localized along with SC-35 in the enlarged rounded speckles and the diffuse staining was not observed (panels e and g). These observations strongly suggest that NPM/B23
phosphorylated on Thr199 function in pre-mRNA processing along with other splicing
factors.
CDK2/cyclin E-mediated phosphorylation on Thr199 of NPM/B23 enhances RNA
binding. Many of the factors that are involved in splicing of mRNA belong to the group of proteins known as Serine/Arginine containing proteins (SR proteins), which possess
RNA-binding activities (42-44). Moreover, it has been demonstrated for certain SR proteins that their RNA-binding activities are modulated by phosphorylation (45,46). It has previously been shown that NPM/B23 also possesses RNA binding activity (2,20).
Based on our finding that phospho-Thr199 NPM specifically localized to nuclear speckles,
we hypothesized that NPM/B23 function in mRNA processing may be controlled by
phosphorylation on Thr199 by CDK2/cyclin E. We thus examined whether Thr199
phosphorylation affects NPM/B23’s ability to bind RNA. GST-NPM/wt as well as GST-
NPM/T199A were subjected to an in vitro kinase assay with CDK2/cyclin E, and then
mixed with total RNA extracted from MSFs. The reaction samples were subjected to
sucrose gradient sedimentation assay (20). GST-NPM/wt alone and GST-NPM/wt with
mock kinase reaction were included as controls. Fractions were collected and examined
84 for the presence of RNA by northern blot analysis using an 18S rRNA probe. The rRNA
molecules were eluted in nearly all the fractions irrespective of the absence or the
presence of NPM/B23 (Fig. 4A). The same fractions were also examined for NPM/B23
by immunoblot analysis. There was no difference in the elution patterns of NPM/B23 in
GST-NPM/wt alone and non-phosphorylated GST-NPM/wt incubated with RNA (Fig.
4B, first and second panels, both of which eluted in the fractions 2-8). The non-
phosphorylatable mutant GST-NPM/T199A after the in vitro kinase reaction with
CDK2/cyclin E showed a similar elution pattern (Fig. 4B, third panel). In contrast, the
phosphorylated GST-NPM/wt was eluted in fractions 2 through 10, demonstrating the
significant increase in the ability to bind RNA (Fig. 4B fourth panel). Thus,
CDK2/cyclin E-mediated phosphorylation on Thr199 greatly enhances the RNA binding
activity of NPM/B23, further confirming phospho-NPM/B23’s role in the mRNA
processing event.
Phosphorylation of NPM/B23 influences its interaction with splicing factors. The co-
localization of phospho-Thr199 NPM to splicing factors at the nuclear speckles raised the
possibility that phospho-Thr199 NPM associates with various splicing factors and/or
spliceosomes. To test this possibility, immunoprecipitations were performed using HeLa
cell lysates with anti-phospho-Thr199 NPM antibody. Since phosphorylation of
NPM/B23 influences its localization and RNA binding activity, we also
immunoprecipitated using NPM/B23 antibody that recognizes both the phosphorylated and non-phosphorylated NPM/B23 in order to test whether phosphorylation of NPM/B23 affects its binding ability to various splicing related proteins. Anti-NPM/B23 antibody immunoprecipitated SR protein SC-35 (Fig. 5A, top panel) but not SR protein SF2/ASF
85 (data not shown). NPM/B23 also immunoprecipitated splicing activator RNPS1 and splicing inhibitor hnRNP I (Fig. 5A, bottom panel). However, NPM/B23 did not bind to spliceosome core component, U1 snRNP70, nor splicing repressor, hnRNP A1 (data not shown). Interestingly, contrary to our localization studies, phospho-Thr199 NPM did not bind to any of the various splicing factors (Fig. 5A). Association of non-phosphorylated
NPM/B23 to splicing factors is specific and phosphorylation of NPM/B23 decreases this association (Fig. 5B).
Phospho-Thr199 NPM represses pre-mRNA splicing. Several kinases, such as Clk/Sty
(47) and SRPK1 (48), have been shown to phosphorylate SR proteins and the phosphorylation and dephosphorylation of SR proteins are required for their splicing acitivity (45-47,49). The observations that phosphorylation of NPM/B23 on Thr199 enhances NPM/B23’s ability to bind RNA and phospho-Thr199 NPM co-localizes with splicing factor SC-35 and splicing activator RNPS1 suggest that phospho-Thr199 NPM may function in pre-mRNA splicing. Therefore, we tested the possible role of phospho-
Thr199 NPM in regulating pre-mRNA splicing activity by in vitro splicing assay. Mock or CDK2/cyclin E phosphorylated GST, GST-NPM/wt or GST-NPM/T199A was added to either HeLa cell nuclear extract or cytosolic S100 extract complemented with SR protein SF2/ASF, which is essential for splicing (50), and subjected to an in vitro splicing assay using β-globin pre-mRNA as substrate (Fig. 6) (33). In HeLa cell nuclear extract splicing assay, mock-phosphorylated GST, GST-NPM/wt, GST-NPM/T199A (Fig. 6, lanes 2, 4 and 6, respectively) showed similar spliced product of β-globin pre-mRNA as well as control, where no GST proteins were added (lane 1). However, in the
CDK2/cyclin E phosphorylated GST-NPM/wt (lane 5), we observed a significant
86 repression of the spliced β-globin pre-mRNA. This splicing repression is specific to the
CDK2/cyclin E mediated phosphorylation on Thr199 of NPM/B23, since CDK2/cyclin E phosphorylated non-phosphorylatable GST-NPM/T199A mutant did not repress splicing
(lane 7).
87 Discussion
NPM/B23 has been implicated in various cellular events including ribosome assembly, intracellular trafficking, DNA polymerase activity, centrosome duplication, and molecular chaperoning. NPM/B23 is phosphorylated by several different kinases on different residues; nuclear kinase II phosphorylates Ser125 (17), CKII phosphorylates
Ser125 and possibly Thr185 (14,16), CDK1/cyclin B phosphorylates residues Thr199,
Thr219, Thr234 and Thr237, and CDK2/cyclin E and CDK2/cyclin A phosphorylates
Thr199 (19,20). Although the functional significance of many of these phosphorylations remains undetermined, these phosphorylations likely affect the biological functions of
NPM/B23. For instance, we have previously shown that CDK2/cyclin E-mediated phosphorylation of NPM/B23 plays an important role in the regulation of centrosome duplication: NPM/B23 changes its affinity to the centrosomes upon CDK2/cyclin E- mediated phosphorylation, and ectopic expression of a dominant negative mutant
NPM/B23 [NPM/T199A] results in inhibition of initiation of centrosome duplication
(11,19). In this study, by use of the antibody that specifically recognizes NPM/B23
phosphorylated on Thr199, we examined whether NPM/B23 shows any specific
subcellular localization upon Thr199 phosphorylation. We found that phospho-Thr199
NPM localized to nuclear speckles as demonstrated by co-localization with the splicing factors SC-35 and RNPS1. The nuclear speckles are thought to function in the storage of splicing factors and to supply splicing factors to the sites of transcription (38,51,52). A family of non-snRNP splicing factors found abundantly in the nuclear speckles, known as
SR proteins, has been studied extensively for its speckle localization signal and its
diverse functions in pre-mRNA splicing. However, the molecular mechanism of how
88 these splicing factors localize to the nuclear speckles is not completely understood. SR proteins, such as SC-35, contain one ribonucleoprotein-type RNA binding domain (RBD)
(53) at the N-terminal, but require the arginine/serine-rich domain (RS domain) for proper targeting to nuclear speckles (54). However, some SR proteins, such as SF2/ASF, contain two RBDs instead of one. It has been shown that the RS domain is not necessary or sufficient for this type of SR proteins to localize to speckles, but the presence of two
RBD domains is critical for speckle localization (54,55). In contrast to the SR proteins,
SAP155 (a subunit of U2 snRNP-associated splicing complex SF3b) contains neither the
RNA binding domain nor RS domain but still localizes to the speckles. This speckle localization is dependent on a highly enriched dipeptide threonine/proline (TP) domain
(56), which are possible CDK phosphorylation sites (57). Sequence analysis of
NPM/B23 shows neither the RS domain nor any apparent RNA binding domain, suggesting that NPM/B23 may localize to the speckles through interacting with other nuclear speckle proteins, or through its RNA binding activity.
Many studies have shown that the phosphorylation and dephosphorylation of SR proteins can determine sub-cellular localization, RNA binding ability, assembly and function of active spliceosomes and regulation of pre-mRNA splicing (58). In this study we found that CDK2/cyclin E mediated phosphorylation of NPM/B23 enhanced its RNA binding activity. We have also found that NPM/B23 binds to certain proteins involved in pre-mRNA splicing; such as SC-35, RNPS1 and hnRNP I. Since the two SR proteins examined in this study (SC-35 and SF2/ASF) are of different sub-types and other splicing factors examined for co-immunoprecipitation (RNSP1, hnRNP I, hnRNP A1) are involved in various mechanisms of pre-mRNA splicing, we concluded that the
89 association of NPM/B23 to the splicing factors is specific. Also this association depends on the phosphorylation status of NPM/B23, as phospho-Thr199 NPM was unable to bind to these factors. Moreover, we found that NPM/B23 plays a role in pre-mRNA splicing.
The phosphorylation of NPM/B23 on Thr199 acts as a repressor of pre-mRNA splicing, while the non-phosphorylatable mutant was able to process pre-mRNA splicing.
The spliceosome is a dynamic multi-protein/RNA complex that undergoes multiple assembly steps and conformational changes during the splicing reaction. Since,
NPM/B23 has molecular chaperone activity, one could speculate that NPM/B23 in its unphosphorylated form acts as a chaperone protein in the association of splicing factors/spliceosome. Upon phosphorylation by CDK2/Cyclin E, phospho-Thr199 NPM dissociates from the components, leading to the disassembly of spliceosome, thus resulting in the repression of pre-mRNA splicing. Once phospho-Thr199 NPM dissociates from the spliceosome, it could be recruited back to the IGCs, where we observed intense phospho-Thr199 NPM localization in the IGCs (Fig. 3). In support of this idea, recently it was shown that phosphorylation of NPM/B23 promoted dissociation of NPM/B23- substrate complex formed in a chaperone activity (16).
Cell cycle-dependent accumulation of phospho-Thr199 NPM at the speckles raises the question of whether some pre-mRNA splicing is controlled in a cell cycle-dependent manner. There is growing evidence linking cell cycle progression with mRNA processing (59,60). A SR protein SRp38 has been shown to act as a splicing repressor during M phase of the cell cycle, during which cell cycle specific dephosphorylation of
SRp38 plays a role in gene silencing (61). Also, it has been shown that the mRNA level of the SRp20 splicing factor is controlled in a cell cycle dependent manner, activated by
90 CDK2/cyclin E-regulated transcription factor E2F (62). Considering that NPM/B23 possesses molecular chaperoning activity, it is possible that NPM/B23 may be involved in splicesome assembly at the site of transcription. In this context, CDK2/cyclin E- mediated phosphorylation of NPM/B23 resulting in accumulation of phospho-Thr199
NPM at the speckles may provide a possible link between cell cycle machinery and pre- mRNA processing.
Acknowledgements- We thank E. Hunter, D. Choi and K. George for their technical assistance, Dr. N. Kleene for quantitative microscopic analysis, and all the members of
Dr. Fukasawa and Dr. Mayeda Labs for helpful discussion. This research is supported by
National Institute of Health (CA90922). Y. Tokuyama is an Albert J. Ryan fellow and a
Robert and EmmaLou Cardell fellow.
91 References
1. Spector, D. L., Ochs, R. L., and Busch, H. (1984) Chromosoma 90, 139-148
2. Wang, D., Baumann, A., Szebeni, A., and Olson, M. O. (1994) J Biol Chem 269,
30994-30998
3. Savkur, R. S., and Olson, M. O. (1998) Nucleic Acids Res 26, 4508-4515
4. Herrera, J. E., Savkur, R., and Olson, M. O. (1995) Nucleic Acids Res 23, 3974-
3979
5. Takemura, M., Sato, K., Nishio, M., Akiyama, T., Umekawa, H., and Yoshida, S.
(1999) J Biochem (Tokyo) 125, 904-909
6. Borer, R. A., Lehner, C. F., Eppenberger, H. M., and Nigg, E. A. (1989) Cell 56,
379-390
7. Adachi, Y., Copeland, T. D., Hatanaka, M., and Oroszlan, S. (1993) J Biol Chem
268, 13930-13934
8. Valdez, B. C., Perlaky, L., Henning, D., Saijo, Y., Chan, P. K., and Busch, H.
(1994) J Biol Chem 269, 23776-23783
9. Szebeni, A., Herrera, J. E., and Olson, M. O. (1995) Biochemistry 34, 8037-8042
10. Szebeni, A., Mehrotra, B., Baumann, A., Adam, S. A., Wingfield, P. T., and
Olson, M. O. (1997) Biochemistry 36, 3941-3949
11. Okuda, M., Horn, H. F., Tarapore, P., Tokuyama, Y., Smulian, A. G., Chan, P. K.,
Knudsen, E. S., Hofmann, I. A., Snyder, J. D., Bove, K. E., and Fukasawa, K.
(2000) Cell 103, 127-140
12. Szebeni, A., and Olson, M. O. (1999) Protein Sci 8, 905-912
92 13. Zhang, H., Shi, X., Paddon, H., Hampong, M., Dai, W., and Pelech, S. (2004) J
Biol Chem
14. Chan, P. K., Aldrich, M., Cook, R. G., and Busch, H. (1986) J Biol Chem 261,
1868-1872
15. Jones, C. E., Busch, H., and Olson, M. O. (1981) Biochim Biophys Acta 667, 209-
212
16. Szebeni, A., Hingorani, K., Negi, S., and Olson, M. O. (2003) J Biol Chem 278,
9107-9115
17. Chan, P. K., Liu, Q. R., and Durban, E. (1990) Biochem J 270, 549-552
18. Peter, M., Nakagawa, J., Doree, M., Labbe, J. C., and Nigg, E. A. (1990) Cell 60,
791-801
19. Tokuyama, Y., Horn, H. F., Kawamura, K., Tarapore, P., and Fukasawa, K.
(2001) J Biol Chem 276, 21529-21537
20. Okuwaki, M., Tsujimoto, M., and Nagata, K. (2002) Mol Biol Cell 13, 2016-2030
21. Koff, A., Giordano, A., Desai, D., Yamashita, K., Harper, J. W., Elledge, S.,
Nishimoto, T., Morgan, D. O., Franza, B. R., and Roberts, J. M. (1992) Science
257, 1689-1694
22. Dulic, V., Lees, E., and Reed, S. I. (1992) Science 257, 1958-1961
23. Spector, D. L. (1993) Annu Rev Cell Biol 9, 265-315
24. Lamond, A. I., and Spector, D. L. (2003) Nat Rev Mol Cell Biol 4, 605-612
25. Mayeda, A., Badolato, J., Kobayashi, R., Zhang, M. Q., Gardiner, E. M., and
Krainer, A. R. (1999) Embo J 18, 4560-4570
93 26. Higuchi, R. (1989) PCR Technology (Erlich, H. A., Ed.), Stockton Press, New
York
27. Desai, D., Wessling, H. C., Fisher, R. P., and Morgan, D. O. (1995) Mol Cell Biol
15, 345-350
28. Sambrook, J., Fritsch, E.F., and Maniatis, T. (1989) Molecular Cloning: A
Laboratory Manual, 2nd Ed., Cold Spring Harbor, New York
29. Seghezzi, W., Chua, K., Shanahan, F., Gozani, O., Reed, R., and Lees, E. (1998)
Mol Cell Biol 18, 4526-4536
30. Mayeda, A., and Krainer, A. R. (1999) Methods Mol Biol 118, 309-314
31. Krainer, A. R., Maniatis, T., Ruskin, B., and Green, M. R. (1984) Cell 36, 993-
1005
32. Mayeda, A., Helfman, D. M., and Krainer, A. R. (1993) Mol Cell Biol 13, 2993-
3001
33. Mayeda, A., and Krainer, A. R. (1999) Methods Mol Biol 118, 315-321
34. Yung, B. Y., Busch, H., and Chan, P. K. (1985) Biochim Biophys Acta 826, 167-
173
35. Fu, X. D., and Maniatis, T. (1990) Nature 343, 437-441
36. Spector, D. L., Fu, X. D., and Maniatis, T. (1991) Embo J 10, 3467-3481
37. Misteli, T. (2000) J Cell Sci 113 ( Pt 11), 1841-1849
38. Jimenez-Garcia, L. F., and Spector, D. L. (1993) Cell 73, 47-59
39. Carmo-Fonseca, M., Pepperkok, R., Carvalho, M. T., and Lamond, A. I. (1992) J
Cell Biol 117, 1-14
94 40. Lindell, T. J., Weinberg, F., Morris, P. W., Roeder, R. G., and Rutter, W. J.
(1970) Science 170, 447-449
41. Spector, D. L. (1996) Exp Cell Res 229, 189-197
42. Fu, X. D. (1995) Rna 1, 663-680
43. Graveley, B. R. (2000) Rna 6, 1197-1211
44. Tacke, R., Chen, Y., and Manley, J. L. (1997) Proc Natl Acad Sci U S A 94, 1148-
1153
45. Xiao, S. H., and Manley, J. L. (1997) Genes Dev 11, 334-344
46. Xiao, S. H., and Manley, J. L. (1998) Embo J 17, 6359-6367
47. Prasad, J., Colwill, K., Pawson, T., and Manley, J. L. (1999) Mol Cell Biol 19,
6991-7000
48. Gui, J. F., Lane, W. S., and Fu, X. D. (1994) Nature 369, 678-682
49. Cao, W., Jamison, S. F., and Garcia-Blanco, M. A. (1997) Rna 3, 1456-1467
50. Krainer, A. R., Conway, G. C., and Kozak, D. (1990) Genes Dev 4, 1158-1171
51. Bridge, E., Xia, D. X., Carmo-Fonseca, M., Cardinali, B., Lamond, A. I., and
Pettersson, U. (1995) J Virol 69, 281-290
52. Misteli, T., Caceres, J. F., and Spector, D. L. (1997) Nature 387, 523-527
53. Burd, C. G., and Dreyfuss, G. (1994) Science 265, 615-621
54. Caceres, J. F., Misteli, T., Screaton, G. R., Spector, D. L., and Krainer, A. R.
(1997) J Cell Biol 138, 225-238
55. Gama-Carvalho, M., Krauss, R. D., Chiang, L., Valcarcel, J., Green, M. R., and
Carmo-Fonseca, M. (1997) J Cell Biol 137, 975-987
95 56. Eilbracht, J., and Schmidt-Zachmann, M. S. (2001) Proc Natl Acad Sci U S A 98,
3849-3854
57. Nigg, E. A. (1995) Bioessays 17, 471-480
58. Misteli, T., Caceres, J. F., Clement, J. Q., Krainer, A. R., Wilkinson, M. F., and
Spector, D. L. (1998) J Cell Biol 143, 297-307
59. Blencowe, B. J. (2003) Curr Biol 13, R149-151
60. Burns, C. G., and Gould, K. L. (1999) Front Horm Res 25, 59-82
61. Shin, C., and Manley, J. L. (2002) Cell 111, 407-417
62. Jumaa, H., Guenet, J. L., and Nielsen, P. J. (1997) Mol Cell Biol 17, 3116-3124
96 FIGURE LEGENDS
Figure 1. Phospho-Thr199 NPM antibody specifically recognizes NPM/B23
phosphorylated on Thr199 by CDK2/cyclin E.
A, Wild-type and mutant NPM/B23 fused to GST (GST-NPM/wt, GST-NPM/T199A) as
well as GST alone was subjected to an in vitro kinase assay with either CKII or
CDK2/cyclin E. The reactions were run on SDS-PAGE and immunoblotted using anti-
phospho-Thr199 NPM antibody (Cell Signaling Technology) (left panel). Coomassie Blue
staining of the same gel is shown on the right. Only the wild-type NPM/B23
phosphorylated by CDK2/cyclin E was detected by the antibody (lane 8).
B, MSFs were fixed and co-immunostained with anti-NPM monoclonal (panel a) and
anti-phospho-Thr199 NPM polyclonal (panel d) antibodies. Antigen-antibody complexes were detected using Alexa Fluor 488-conjugated goat anti-mouse IgG (for anti-NPM monoclonal antibody) or goat anti-rabbit IgG antibodies (for anti-phospho-Thr199 NPM polyclonal antibody) (Molecular Probes). The nucleoli are depicted by differential interference contrast microscopy (panels b and e). Panels c and f show the overlay images of panels a and b and panels d and e, respectively. Scale bar; 10 µm.
C, MSFs treated with alkaline phosphatase (panels a and b) and mock-treated MSFs
(panels c and d) were immunostained with anti-phospho-Thr199 NPM polyclonal
antibody. Antigen-antibody complexes were detected using Alexa Fluor 488 goat anti-
rabbit IgG antibody (panels a and c). The nuclei were also counterstained with DAPI
(panels b and d). In control cells, phospho-Thr199 NPM was distributed in a speckled pattern (panel a). In contrast, in the alkaline phosphatase treated cells, anti-phospho-
97 Thr199 NPM antibody no longer detected the speckled pattern (panel c). Scale bar; 10
µm.
Figure 2. Cell-cycle dependent changes in the level of phospho-Thr199 NPM.
A, MSFs were serum starved for 48 h, followed by serum stimulation with medium
containing 20% FBS. At indicated time points, cells were immunostained using anti-
phospho-Thr199 NPM polyclonal antibody (0h-20h). B, The immunostaining images
taken by confocal microscopy for (A) were quantitatively analyzed for phospho-Thr199
NPM signals by integrated morphometry analysis using Metamorph software program.
In parallel, MSFs were examined for BrdU incorporation (C). After serum starvation for
48h, MSFs were serum stimulated with medium containing 20% FBS in the presence of
BrdU. MSFs were then immunostained with anti-BrdU monoclonal antibody. The percent of BrdU incorporated cells was determined by examining >300 cells by fluorescence microscopy. D, MSFs were serum starved for 48h, followed by serum stimulation. The whole cell lysates were prepared at indicated time points and immunoblotted with anti-phospho-Thr199 NPM polyclonal antibody (left panel) as well as
anti-NPM monoclonal antibody (middle panel). The right panel shows the Commassie
blue-stained membrane.
Figure 3. Phospho-Thr199 NPM co-localizes with splicing factors and re-distributes
upon transcription inhibition.
A, Exponentially growing MSFs were co-immunostained with anti-phospho-Thr199 NPM
polyclonal (panel a) and anti-splicing factor SC-35 monoclonal (Sigma-Aldrich) (panel
b) antibodies. Antigen-antibody complexes were detected with Alexa Fluor 488 goat
anti-rabbit IgG (green) and Alexa Fluor 594 goat anti-mouse IgG (red) antibodies. Panel
98 c is the overlay image, showing a high degree of co-localization of phospho-Thr199 NPM and SC-35. B, MSFs were transiently transfected with plasmids expressing T7-tagged
RNPS1 and co-immunostained with anti-phospho-Thr199 NPM polyclonal (panel a) and
anti-T7 tag monoclonal (Novagen) (panel b) antibodies. Antigen-antibody complexes
were detected as described in (A). The co-localization of phospho-Thr199 NPM and
RNPS1 was detected as yellow color in the overlay image (panel c). Scale bar; 10 µm.
C, Exponentially growing MSFs were treated with α-amanitin (50 µg/ml) for 5 h (panels
e-h). The control parallel culture of MSFs were not treated with α-amanitin (panels a-d).
Cells were then fixed for immunostaining with anti-phospho-Thr199 NPM polyclonal and
anti-SC-35 monoclonal antibodies as described in the legend to Fig. 4. The nuclei were
also counterstained with DAPI (panels d and h). Panels c and g are the overlay images of
panels a and b and panels e and f, respectively. In control cells, phospho-Thr199 NPM
(panel a) and SC-35 (panel b) co-distributed in a speckled pattern along with a diffuse staining. In contrast, in α-amanitin treated cells, both phospho-Thr199 NPM and SC-35
were redistributed to the rounded and enlarged speckles. Scale bar; 10 µm.
Figure 4. Phosphorylation on Thr199 enhances the RNA-binding activity of
NPM/B23.
GST-NPM/wt and GST-NPM/T199A were subjected to an in vitro kinase assay with
CDK2/cyclin E and incubated in the absence or presence of total RNA prepared from
MSFs. The reaction samples were then subjected to a 15-40% sucrose gradient
fractionation, and the fractions were collected from the bottom of the tube. RNA as well
as RNA/protein complexes were precipitated from each fraction. RNA from each
fraction was resolved by 1% agarose-formaldehyde gel, probed for 18S rRNA (Ambion)
99 as described in Materials and Methods. The intensity of each band from the Northernblot was calculated using Metamorph software (A). Each fraction was also resolved by 10%
SDS-PAGE, and was subjected to immunoblot analysis using anti-NPM monoclonal antibody (B, bottom four panels). The arrows point to the fractions showing apparent differences in the RNA-binding affinity of NPM/B23 upon CDK2/cyclin E-mediated phosphorylation.
Figure 5. Phosphorylation on Thr 199 influences association of NPM/B23 with splicing factors.
A, Immunoprecipitation from HeLa nuclear extracts were performed with anti-NPM monoclonal and anti-phospho-Thr199 NPM polyclonal antibodies. Representative blots of immunoprecipitates are shown that were immunoblotted with SC-35 and hnRNPI antibodies. B, Summary table of immunoprecipitation from (A) immunoblotted with various splicing related factors: SC-35 (SR protein), SF2/ASF (SR protein), hnRNPI
(splicing inhibitor), hnRNPA1 (splicing repressor), U1 snRNP70 (component of spliceosome) and RNPS1 (splicing activator).
Figure 6. Phosphorylation on Thr199 represses splicing in vitro.
β-globin pre-mRNA was incubated with constant amounts of HeLa nuclear extracts with mock (M) or CDK2/cyclin E phosphorylated (P) GST, GST-NPM/wt, GST-NPM/T199A in the splicing reaction.
100 A Control CKII CDK2/CycE Control CKII CDK2/CycE
A A A A A 9 A 9 9 9 9 9 9 9 9 9 9 T 1 T T 9 T T T 1 1 1 1 1 T T W / W T W W W T W T / / / / T / / / / / / / M M M M M M M M M M M P M P P P P P P P P P P P N N - N N N N N N N N N N ------T T T T T T T T T T T T T T T T T S T S S S S S S S S S S S S S S S S S G G G G G G G G kDa G G G G G G G G G G 150 100 75 GST-NPM 50
37
25 GST 123 456 789 10 11 12 13 14 15 16 17 18 Phospho-Thr199NPM Commassie Blue
B NPM DIC (nucleolus) Overlay ab c
Anti-NPM Ab
de f
Anti-Phospho-Thr199 NPM Ab
C Phospho-Thr199 NPM DAPI a b
control
cd
AP-treated Figure 1
101 A
0h 5h 10h
15h 20h
BC
25000 100 ) % 20000 80 NPM
199 15000 60 ated cells ( Thr por 10000 40 ncor i - hospho- integrated intensity P 5000 dU 20 Br
0 0 0 5 10 15 20 0 5 10 15 20 Time after serum-stimulation (hr) Time after serum-stimulation (hr) D
kDa 0 5 10 15 20 h 0 5 10 15 20 h 0 5 10 15 20 h 75
50
199 Total 37 Thr NPM NPM
WB: Phospho-Thr 199NPM Ab WB: NPM Ab Commassie Blue
Figure 2
102 A
Phospho-Thr 199 NPM SC-35 Overlay acb
B Phospho-Thr 199 NPM α - T7 (RNPS1) Overlay acb
C Phospho-Thr 199 NPM SC-35 Overlay DAPI acb d
e fgh
Figure 3
103 A
Fr 12 3 485 697 10
NPM/wt + CDK2/cyclin E + RNA NB:18S rRNA
B Fr 123485697 10
NPM/wt
NPM/wt + RNA WB:NPM Ab NPM/T199A + CDK2/cyclin E + RNA
NPM/wt + CDK2/cyclin E + RNA
Figure 4
104 A
5% Phospho- Input IP: NPM IgG Thr199 NPM IgG
SC-35
hnRNPI
B
U1 SC-35 SF2/ASF hnRNPI hnRNPA1 RNPS 1 WB snRNP70
NPM + - + - - + IP Phospho------Thr199 NPM
Figure 5
105
Nuclear Extract
M P - - M P - - M A A - P 9 T T - 9 9 9 S S T T 1 1 G G W W T T
pre-mpre-mRNA RNA
mRNA
QuickTime™ and a TIFF (Uncompressed) decompressor are needed to see this picture.
123 4 567
Figure 6
106
Chapter IV
Ubiquitination of NPM/B23
Yukari Tokuyama and Kenji Fukasawa
Department of Cell Biology, University of Cincinnati College of Medicine PO Box 670521, Cincinnati, OH 45267-0521
107 Abstract
The centrosomes direct the formation of bipolar mitotic spindles, which are essential for accurate chromosome transmission to daughter cells. Recent studies have shown that centrosome amplification is commonly observed, and is the major contributing factor for chromosome instability in human cancers. Centrosome duplication is regulated by the kinase activity of cyclin-dependent kinase 2 (CDK2). Previously we have found nucleophosmin (NPM/B23) as a substrate of CDK2/cyclin E in the initiation of centrosome duplication. NPM/B23 associates with single (unduplicated) centrosomes and upon phosphorylation by CDK2/cyclin E on Thr199 amino acid residue, NPM/B23 is lost from the centrosomes, resulting in the initiation of centrosome duplication. Here we have found that NPM/B23 undergoes another post-translational modification: ubiquitination. Moreover, the phosphorylation on Thr199 is required for its ubiquitination and this could be a possible mechanism for NPM/B23’s regulation at the centrosome.
108 Introduction
The centrosome, a microtubule-organizing center, consists of a pair of centrioles surrounded by numerous proteins termed pericentriolar materials (PCMs). During mitosis, the centrosomes direct the formation of bipolar mitotic spindles, which are essential for accurate chromosome segregation to each daughter cell. Since each daughter cell receives one centrosome, the centrosome must duplicate once and only once prior to the next mitosis. Thus, centrosome duplication is coordinated with other cell cycle events (i.e. DNA synthesis) (1). In mammalian cells, centrosome duplication is initiated upon activation of cyclin-dependent kinase 2 (CDK2) at G1/S transition (2-5).
The activation of CDK2 is regulated by the maximal expression of cyclin E in late G1 and active CDK2/cyclin E initiates both centrosome duplication and DNA replication.
Constitutive activation of CDK2/cyclin E has been shown to disrupt the coordination of centrosome duplication and DNA replication, where centrosome duplicates immediately after G1 phase entry (6). Thus, disruption of this regulation of coordination of centrosome duplication and DNA replication increases the frequency of centrosome amplification. Studies show centrosome amplification contributes to genetic instability which is a hallmark for many human tumors (7).
We have identified nucleophosmin (NPM/B23) as a substrate of CDK2/cyclin E in the initiation of centrosome duplication (8). NPM/B23 (also known as numatrin,
NO38) is a phosphoprotein abundant in the nucleolus. It is thought to play a role in ribosome biogenesis from its localization and association with pre-ribosomal particles
(9,10), and its intrinsic ribonuclease activity (11). NPM/B23 has also been implicated in nucleocytoplasmic shuttling (12-14), stimulation of DNA polymerase α (15) and
109 molecular chaperone activities (16,17). Previously, we have shown that NPM/B23
associates with unduplicated centrosomes and upon phosphorylation by CDK2/cyclin E
on Thr199, NPM/B23 is lost from the centrosome. When the non-phosphorylatable
NPM/B23 mutant (Thr199→Ala) is expressed in cells, it acts as a dominant negative
resulting in the suppression of centrosome duplication. These studies suggest that
CDK2/cyclin E phosphorylation on Thr199 of NPM/B23 is essential for loss of
centrosomal NPM/B23 and initiating centrosome duplication (8,18).
Recently, it has been shown that ubiquitin-mediated proteolysis plays an
important role in centrosome duplication (19). Also, studies have shown that active
proteasomal complexes and ubiquitin are localized at the centrosome (20,21). In
Saccharomyces cerevisiae, separation of spindle pole bodies requires components of one
of many E3 ligases, SCF (Skp1-cullin-F-box) ubiquitin ligase (22). In mammalian cells,
Skp1 and Cul1 (components of the SCF ubiquitin ligase) have been shown to localize to
the centrosomes (23,24). Also, centrosome localized Cul1 is modified by Nedd8,
ubiquitin-like molecule, which is necessary for SCF ubiquitin activity. Moreover,
antibody injection study of Skp1 and Cul1 has been shown to inhibit centriole separation
that could not be restored by addition of CDK2/cyclin E (23). This suggests that SCF-
mediated proteolysis is targeting a protein other than cyclin dependent kinase inhibitor, which led us to question whether NPM/B23 could be the target protein.
In this study, we show that proteasome inhibitor prevents loss of NPM/B23 from the centrosome and centrosome duplication is suppressed. Also, NPM/B23 is maximally ubiquitinated at G1/S phase transition in vivo. Moreover, the phosphorylation on Thr199
110 by CDK2/cyclin E is required for its ubiquitination in vitro. Thus, ubiquitin modification of NPM/B23 may be the key event in initiating centrosome duplication.
111 Materials and Methods
Cell Culture.
HeLa cells and mouse skin fibroblasts (MSFs) were maintained in complete medium
(Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum,
penicillin (100 units/ml), and streptomycin (100 µg/ml)) in an atmosphere containing
10% CO2.
Immunoblot Analysis.
Samples were denatured at 95 °C for 5 min in sample buffer (2% SDS, 10% glycerol,
60mM Tris (pH6.8), 5% β-mercaptoethanol, 0.01% bromphenol blue). Samples were
resolved by SDS-PAGE and transferred onto Immobilon-P (Millipore Corp.) sheets. The
blots were first incubated in blocking buffer (5% (w/v) nonfat dry milk in Tris-buffered
saline plus Tween 20 (TBS-T)) for 1 h. Then the blots were incubated with primary
antibody overnight at 4 °C, followed by incubation with horseradish peroxidase- conjugated secondary antibody for 1 h at room temperature. Samples were washed five times with TBS-T after each incubation. The antibody-antigen complex was visualized by ECL chemiluminescence (Amersham Pharmacia Biotech).
Indirect Immunofluorescence.
Cells grown on coverslips were fixed with methanol:acetone (50:50) for 20 min at -20
°C, and then washed with methanol followed by washes with phosphate-buffered saline
(PBS). Cells were incubated with blocking solution (15% normal goat serum in PBS) for
1 h and probed with primary antibodies for 30 min, and antibody-antigen complexes were detected with either rhodamine- or FITC-conjugated goat secondary antibody by incubation for 30 min at room temperature. The samples were washed three times with
112 PBS after each incubation and then counterstained with 4’, 6-diamidino-2-phenylindole
(DAPI).
BrdU Incorporation Assay.
The assay was performed using the BrdUrd labeling kit (Roche Molecular Biochemicals) according to the manufacturer’s instructions. Briefly, cells were fixed in 70% ethanol in
50mM glycine (pH2) for 20 min at -20 °C, incubated in blocking buffer for 1 h at room temperature, and then probed with anti-BrdUrd monoclonal antibody for 30 min at 37°C.
Antigen-antibody complexes were detected by FITC-conjugated sheep anti-mouse IgG.
In Vitro Ubiquitination assay.
GST, GST-NPM/wt or GST-NPM/T199A conjugated with GST-Sepharose 4 beads was
added to ubiquitin buffer (50mM Tris (pH 7.6), 2mM MgCl2, 1mM DTT), 25µg
cytoplasmic extracts prepared from exponentially growing HeLa cells by hypotonic
buffer (20mM Hepes (pH7.2), 5mM KCl, 1.5mM MgCl2, 0.5mM DTT), 1µM ubiquitin
aldehyde (BostonBiochem), 1mg/ml methylated ubiquitin (BostonBiochem) and 2mM
ATP. The reaction mixture was then incubated for 2 hr at 30 °C. Following incubation, the substrates were centrifuged and washed with PBS. Then sample loading buffer was added to the substrates and boiled for 5 min and analyzed by immunoblotting with anti- ubiquitin polyclonal antibody (Sigma).
113 Results
Proteasome Inhibitor Suppresses Centrosome Duplication. Studies have shown that
proteasome function may be required for SPB (spindle pole body) and centrosome
duplication and separation (23,25). To test the involvement of proteasome-mediated
proteolysis in regulation of centrosome duplication, we treated HeLa cells with
proteasome inhibitor clastolactacystin β-lactone (CLBL) and examined the cells for
centrosomes by immunostaining for γ-tubulin, a component of pericentrial material of the
centrosome. In DMSO control cells, ~ 55 % of cells contained single centrosomes and ~
40 % of cells contained duplicated centrosomes (Fig 1A). In contrast, CLBL treated cells contained ~ 75 % of cells with single centrosomes (Fig 1A). To eliminate the possibility that inhibition of centrosome duplication is due to inhibition of cell cycle, the same
CLBL untreated and treated cells were treated with BrdU for 3 hr and immunostained with anti-BrdU monoclonal antibody to examine the cell cycle progression (Fig 1B).
There were no differences in the incorporation of BrdU between untreated and treated cells, indicating that proteasome-dependent proteolysis specifically suppresses centrosome duplication.
Proteasome Inhibitor Prevents Loss of NPM from Centrosomes. Freed et.al. have
shown components of the E3 ubiquitin ligase (SCF complex) to localize at the
centrosome and ubiquitin-mediated proteolysis to be important for centrosome
duplication. Previously, we have identified NPM/B23 as a substrate of CDK2/cyclin E in
initiation of centrosome duplication (8). Thus, to determine whether the proteasome
inhibitor mediated suppression of centrosome duplication is due to inhibition of loss of
NPM/B23 from the centrosome, we examined the association of NPM/B23 with cells
114 containing single centrosomes by co-immunostaining for γ-tubulin and NPM/B23 (Fig 2).
The graph in Fig. 2A shows the profile of NPM/B23 association with single centrosomes.
In DMSO control cells, we observed association of NPM/B23 with ~ 30% of cells
containing single centrosomes. However, in CLBL treated cells NPM/B23 associated
with a majority (~80%) of cells containing single centrosomes. Thus, proteasome
inhibition results in an increase in the association of NPM/B23 with centrosomes,
suggesting that NPM/B23 could be degraded at the centrosomes.
NPM/B23 is ubiquitinated during G1 phase. To determine the involvement of ubiquitin-mediated proteolysis on NPM/B23 in vivo, mouse skin fibroblasts (MSFs) were synchronized by serum-starvation (1% FBS), followed by serum stimulation (20% FBS).
At 0, 4, 8, 12, 16 hr after serum stimulation, cells were harvested and immunoprecipitated in the presence of CLBL for ubiquitin with anti-ubiquitin polyclonal antibody (Sigma), then ubiquitin immunoprecipitates were immunoblotted for NPM/B23 with anti-
NPM/B23 monoclonal antibody. At 0, 4, 8 hr after serum stimulation, low levels of poly- ubiquitinated NPM/B23 (high-molecular weight NPM/B23) was detected (Fig 3A, brackets). However, at 12 hr, we observed high levels of poly-ubiquitination of
NPM/B23 and at 16 hr, the poly-ubiquitination of NPM/B23 was reduced. Also, to examine the cell cycle progression of these MSFs after serum stimulation, same cells were subjected to a BrdU incorporation assay (Fig 3B). After 8 hr of serum stimulation,
BrdU incorporated cells increased, and at 16 hr, ~ 10 % of cells were BrdU positive.
Thus, NPM/B23 is ubiqutinated throughout G1 phase and is maximally poly- ubiquitinated during G1/S transition of the cell cycle.
115 NPM/B23 is ubiquitinated in a phosphorylation dependent manner in vitro.
Ubiquitination of substrate proteins is regulated by post-translational modifications, such as phosphorylation, as the phosphorylation creates binding sites for E3 ligases on the substrates (26). Since phosphorylation of NPM/B23 on residue Thr 199 by CDK2/cyclin
E results in the loss of NPM/B23 from the centrosomes we examined whether this phosphorylation of NPM/B23 on Thr 199 is required for its ubiquitination. Wild-type and non-phosphorylatable mutant NPM/B23 was fused to GST (GST-NPM/wt, GST-
NPM/T199A, respectively) and subjected to an in vitro ubiquitination assay with HeLa cell extracts. Poly-ubiquitination NPM/B23 was readily detected in the GST-NPM/wt
(Fig 4, brackets), whereas no ubiquitination was observed in the non-phosphorylatable
GST-NPM/T199A mutant. The finding that GST-NPM/wt is ubiquitinated but not GST-
NPM/T199A mutant implies that upon phosphorylation by CDK2/cyclin E on Thr 199,
NPM/B23 is targeted for ubiquitination.
116 Discussion
Regulation of the coordination of centrosome duplication and DNA replication is essential to ensure proper chromosome segregation to each daughter cell. Several studies have shown that activation of CDK2 is required for regulation of centrosome duplication.
Previously, we have identified NPM/B23 as a target of CDK2/cyclin E in initiation of centrosome duplication (8). NPM/B23 is regulated at the centrosome by CDK2/cyclin E- mediated phosphorylation, as NPM/B23 is associated with single centrosome and upon phosphorylation on Thr199 by CDK2/cyclin E, NPM/B23 is lost from the centrosome
(18). In this study, we further examined the role of Thr199 phosphorylation on
NPM/B23. We found NPM/B23 is ubiquitinated during G1 phase of cell cycle.
Ubiquitination of NPM/B23 was maximal at the G1/S transition and its ubiquitination was reduced as the cells progressed to S phase. Furthermore, NPM/B23 is ubiquitinated in a CDK2/cyclin E phosphorylation dependent manner as the non-phosphorylatable mutant of NPM/B23 (NPM/T199A) could not be ubiquitinated.
Although, we have not determined the significance of this ubiquitination on
NPM/B23, this study raises several interesting possibilities. There are growing evidence suggesting proteasome-mediated degradation plays a role in centrosome duplication. The
SCF (Skp1-cullin-F-box) ubiquitin ligase has been shown to localize at the centrosome and is thought to initiate the centrosome duplication (23). However, the exact mechanism and substrate of this ubiquitin-mediated degradation is not known. Thus, it is possible that NPM/B23 is the substrate of SCF ubiquitin ligase at the centrosome to initiate centrosome duplication.
117 Also, recently BRCA1-BARD1 complex, which contains an ubiquitin ligase activity, was shown to catalyze Lys6 poly-ubiquitination of NPM/B23 (27). BRCA1 and
BARD1 have been shown to localize at the centrosome during mitosis and throughout the interphase (observations in Fukasawa lab), thus it is possible that BRCA1-BARD1 is responsible for the CDK2/Cyclin E phosphorylation dependent ubiquitination of
NPM/B23. This raises a possibility that ubiquitination of NPM/B23 observed in this study is through a Lys6 linked poly-ubiquitin chain, which signals for a non-proteasomal pathway. Unlike other ubiquitin moieties, very little is known about the conjugation of
Lys6 ubiquitin. Lys6 ubiquitination has been observed in E3-indendpent reactions catalyzed by Rad6, but the function of this Lys6 ubiquitination is not known (28). It is thought that BRCA1-BARD1 ubiquitination of NPM/B23 supports the stabilization of
NPM/B23 rather than its degradation (27). Thus, it is possible that once NPM/B23 is phosphorylated, the poly-ubiquitinated NPM/B23 undergoes a conformational change, which could lead to the loss of NPM/B23 from centrosomes but still stabilizing
NPM/B23 instead of targeting it to a proteasome-mediated pathway.
Another interesting study has demonstrated that tumor suppressor ARF ubiquitinates NPM/B23 for proteasome-mediated degradation (29). Therefore, it is possible that NPM/B23 undergoes different/distinct ubiquitination in which it specifies the fate of NPM/B23 in its biological functions. Further studies, examining the precise ubiquitin moieties that is responsible for centrosomal NPM/B23’s ubiquitination will provide new insight into the role of ubiquitination of NPM/B23 in centrosome duplication.
118 References
1. Hinchcliffe, E. H., and Sluder, G. (2001) Genes Dev 15, 1167-1181
2. Hinchcliffe, E. H., Li, C., Thompson, E. A., Maller, J. L., and Sluder, G. (1999)
Science 283, 851-854
3. Lacey, K. R., Jackson, P. K., and Stearns, T. (1999) Proc Natl Acad Sci U S A 96,
2817-2822
4. Matsumoto, Y., Hayashi, K., and Nishida, E. (1999) Curr Biol 9, 429-432
5. Tarapore, P., Okuda, M., and Fukasawa, K. (2002) Cell Cycle 1, 75-81
6. Mussman, J. G., Horn, H. F., Carroll, P. E., Okuda, M., Tarapore, P., Donehower,
L. A., and Fukasawa, K. (2000) Oncogene 19, 1635-1646
7. Lingle, W. L., and Salisbury, J. L. (2000) Curr Top Dev Biol 49, 313-329
8. Okuda, M., Horn, H. F., Tarapore, P., Tokuyama, Y., Smulian, A. G., Chan, P. K.,
Knudsen, E. S., Hofmann, I. A., Snyder, J. D., Bove, K. E., and Fukasawa, K.
(2000) Cell 103, 127-140
9. Prestayko, A. W., Klomp, G. R., Schmoll, D. J., and Busch, H. (1974)
Biochemistry 13, 1945-1951
10. Yung, B. Y., Busch, H., and Chan, P. K. (1985) Biochim Biophys Acta 826, 167-
173
11. Savkur, R. S., and Olson, M. O. (1998) Nucleic Acids Res 26, 4508-4515
12. Borer, R. A., Lehner, C. F., Eppenberger, H. M., and Nigg, E. A. (1989) Cell 56,
379-390
13. Adachi, Y., Copeland, T. D., Hatanaka, M., and Oroszlan, S. (1993) J Biol Chem
268, 13930-13934
119 14. Valdez, B. C., Perlaky, L., Henning, D., Saijo, Y., Chan, P. K., and Busch, H.
(1994) J Biol Chem 269, 23776-23783
15. Takemura, M., Sato, K., Nishio, M., Akiyama, T., Umekawa, H., and Yoshida, S.
(1999) J Biochem (Tokyo) 125, 904-909
16. Szebeni, A., Hingorani, K., Negi, S., and Olson, M. O. (2003) J Biol Chem 278,
9107-9115
17. Szebeni, A., and Olson, M. O. (1999) Protein Sci 8, 905-912
18. Tokuyama, Y., Horn, H. F., Kawamura, K., Tarapore, P., and Fukasawa, K.
(2001) J Biol Chem 276, 21529-21537
19. Hansen, D. V., Hsu, J. Y., Kaiser, B. K., Jackson, P. K., and Eldridge, A. G.
(2002) Oncogene 21, 6209-6221
20. Fabunmi, R. P., Wigley, W. C., Thomas, P. J., and DeMartino, G. N. (2000) J
Biol Chem 275, 409-413
21. Wigley, W. C., Fabunmi, R. P., Lee, M. G., Marino, C. R., Muallem, S.,
DeMartino, G. N., and Thomas, P. J. (1999) J Cell Biol 145, 481-490
22. Patton, E. E., Willems, A. R., and Tyers, M. (1998) Trends Genet 14, 236-243
23. Freed, E., Lacey, K. R., Huie, P., Lyapina, S. A., Deshaies, R. J., Stearns, T., and
Jackson, P. K. (1999) Genes Dev 13, 2242-2257
24. Gstaiger, M., Marti, A., and Krek, W. (1999) Exp Cell Res 247, 554-562
25. McDonald, H. B., and Byers, B. (1997) J Cell Biol 137, 539-553
26. Fang, S., and Weissman, A. M. (2004) Cell Mol Life Sci 61, 1546-1561
27. Sato, K., Hayami, R., Wu, W., Nishikawa, T., Nishikawa, H., Okuda, Y., Ogata,
H., Fukuda, M., and Ohta, T. (2004) J Biol Chem 279, 30919-30922
120 28. Baboshina, O. V., and Haas, A. L. (1996) J Biol Chem 271, 2823-2831
29. Itahana, K., Bhat, K. P., Jin, A., Itahana, Y., Hawke, D., Kobayashi, R., and
Zhang, Y. (2003) Mol Cell 12, 1151-1164
121 FIGURE LEGENDS
Figure 1. Proteasome inhibitor blocks centrosome duplication.
A. HeLa cells were treated with proteasome inhibitor clasto-lactacystin β-lactone (0,
5µM) for 24 hrs and immunostained for centrosomes by anti-γ-tubulin antibody. The
number of centrosomes per cells (n) were then statistically scored (n=1, n=2, n≥3).
B. During the last 3 hrs of cultering, BrdUrd was added to the medium. The cells were
then processed for immunostaining with anti-BrdUrd antibody and the percentage of cells
that had incorporated BrdUrd was determined.
Figure 2. Proteasome inhibition prevents loss of centrosomal NPM/B23.
HeLa cells were cultured in the presence and absence of 5 uM proteasome inhibitor
clastolactacystin β-lactone (CLBL) for 24 hs. Cells were then fixed with
methanol/acetone (50:50), and co-immunostained for γ-tubulin (B, indicated by
arrows in panels a & e) and NPM/B23 (B, panels b & f) . The nuclei were stained with
DAPI. Panels d & h show the overlay images of γ-tubulin and NPM/B23 staining.
Association of NPM/B23 with single centrosomes were examined (A).
Figure 3. NPM/B23 is ubiquitinated during G1 phase in vivo.
A. Mouse skin fibroblasts (MSFs) were synchronized by serum-starvation (1% FBS), followed by serum stimulation (20% FBS). At 0, 4, 8, 12, 16 hr after serum stimulation, cells were harvested and immunoprecipitated with anti-ubiquitin antibody.
Immunoprecipitated complex was immunoblotted for NPM/B23 to analyze for ubiquitinated NPM/B23.
B. After serum starvation for 48h, MSFs were serum stimulated with medium containing
20% FBS in the presence of BrdU. MSFs were then immunostained with anti-BrdU
122 monoclonal antibody. The percent of BrdU incorporated cells was determined by examining >300 cells by fluorescence microscopy.
Figure 4. In vitro ubiquitination of NPM/B23.
In vitro ubiquitination assay was performed using GST fused wild-type NPM/B23 and non-phosphorylatable mutant (GST-NPM/wt, GST-NPM/T199A, respectively) in the presence of HeLa cell extracts supplemented with methyl-ubiquitin (UbM) and ubiquitin aldehyde (UbA). In the presence of modified ubiquitins (UbM and UbA), accumulation of poly-ubiquitinated GST-NPM/wt was observed as high-molecular weight proteins.
123 A one centrosome two centrosomes > two centrosomes 100
80
60 Percent of cells . 40
20
0 0µM 5µM Proteasome Inhibitor
B
BrdU (+) BrdU (-)
0µM 41.65 % 57.85 %
5µM 45.35 % 54.65 %
Figure 1
124 A NPM association with one centrosome
100
80
60 Percent of cells 40
20
0 0µM 5µM
Proteasome Inhibitor
B
a γ-tubulin b NPM e γ-tubulin f NPM
c DAPI d Overlay g DAPI h Overlay
Figure 2
125
A B
0481216h 20
15
Ub-NPM 10 dU-positive cells IgG r B f o
% 5 NPM
0 048112 6 h
Figure 3
126 UbA+ UbM +Extract
A 9 9 t 1 w T
Ub-NPM
Figure 4
127
Chapter V
Summary and Conclusion
128 Summary
Based upon the studies described here, we have elucidated several novel functions of
NPM/B23. In Chapter II, we have identified a specific phosphorylation site of NPM/B23
by CDK2/cyclin E and its role in centrosome duplication cycle (1). NPM/B23 is
associated with single centrosomes, and upon phosphorylation by CDK2/Cyclin E on
Thr199, NPM/B23 is lost from the centrosomes. Furthermore, the Thr199Ala
phosphorylation mutant acts as a dominant negative, suppressing centrosome duplication.
Thus, NPM/B23 may act as a licensing factor for centrosome duplication. Further
examination of the role of CDK2/Cyclin E-mediated phosphorylation on NPM/B23
revealed localization of phospho-NPM/B23 to specific nuclear domain and a function in
pre-mRNA splicing, which is demonstrated in Chapter III. NPM/B23 phosphorylated on
Thr199 by CDK2/cyclin E localizes to nuclear speckles. NPM/B23 is associated with several specific splicing factors and upon phosphorylation of NPM/B23; this association is decreased. Moreover, phosphorylation of NPM/B23 on Thr199 represses pre-mRNA splicing. Since the phosphorylation of NPM/B23 is cell cycle dependent, these finding suggests a possible link between cell cycle progression and pre-mRNA splicing. Chapter
IV examines the fate of phosphorylation of NPM/B23 on Thr199 by CDK2/Cyclin E.
We have found NPM/B23 to undergo ubiquitination in a phosphorylation-dependent manner. The exact role of this ubiquitination on NPM/B23’s function and stability is not clear, but possible mechanisms will be discussed in this chapter.
NPM/B23 and centrosome duplication
129 The structure and function of the centrosome has been studied since its discovery
in the late 1800’s by Edouard van Beneden and Theodor Boveri. In 1914, Boveri proposed that tumor cells could arise by the aberrant replication and activity of centrosomes, producing multiple spindle poles that lead to aberrant chromosome distribution between daughter cells; resulting in the gain and/or loss of chromosomes
(aneuploidy) during mitosis (2). However, it was not until recently that Boveri’s proposition of centrosomes’ role in tumor progression came to light. In order to understand the mechanism of chromosome instability observed in mice deficient for the p53 tumor suppressor, Fukasawa et al., found that p53 regulates the duplication of centrosomes, and showed that loss of p53 resulted in abnormal amplification of centrosomes (3,4). Since then, the study of centrosomes has attracted many investigators, especially in regard to understanding the regulation of the centrosome duplication cycle and the role of the centrosome in chromosome instability. Recently, we have identified
NPM/B23 as a target protein in the initiation of centrosome duplication (5). As described in Chapter II, NPM/B23 is associated with single centrosomes and is lost from the centrosomes once it is phosphorylated by CDK2/Cyclin E, thus acting as a licensing factor in the centrosome duplication cycle. However, many questions still remain to be answered. To further our understanding regarding the mechanism underlying
NPM/B23’s function at the centrosome, it will be important to examine the precise timing of the association/localization of NPM/B23 at the centrosome and its exact position on the centrioles. In a preliminary study, we developed a new NPM/B23 antibody that recognizes a different epitope (N-terminus of NPM/B23) than our original NPM/B23 antibody (C-terminus of NPM/B23). The new NPM/B23 antibody was used to co-
130 immunostain for NPM/B23 and centrioles, and we demonstrated that NPM/B23 localizes in between the two centrioles (Fig.1).
NPM NPM centriole
centriole centriole centriole
Figure 1. 3D-deconvolution analysis of NPM/B23 at the centrosome.
MSFs were fixed and co-immunostained with anti-NPM/B23 (green) and
anti-centrin (red) antibodies. Z-stack images were obtained and then
subjected to a 3D-deconvolution analysis using Metamorph software.
Our studies using the non-phosphorylatable mutant of NPM/B23 indicated that it inhibited centrosome duplication at an early stage during the duplication cycle, and that it was bound to the centriole pair with an intact orthogonal configuration (1,5). These observations suggest that NPM/B23 may function as a connector between the pair of centrioles. Another interesting question is what happens to the centrosomal phosphorylated NPM/B23. We did not observe NPM/B23 at the centrosomes once the centrosomes duplicated; however, NPM/B23 re-associated with the centrosomes during mitosis (5,6). It is possible that ubiquitination of NPM/B23 plays a role in the regulation of NPM/B23 localization at the centrosomes. Since the ubiquitin-mediated proteasome
131 complex is shown to localize to the centrosome (7-9), we examined the possibility of ubiquitin-mediated degradation of NPM/B23, which is discussed in Chapter IV. We found that phosphorylation of NPM/B23 by CDK2/Cyclin E targets NPM/B23 for ubiquitination. Although we have not determined the result of the ubiquitination on
NPM/B23, it is possible that ubiquitination targets NPM/B23 for degradation.
Many questions remain in regard to whether the ubiquitination of NPM/B23 plays a role in centrosome duplication. Since there is growing evidence that ubiquitination is not just a signal for degradation (10,11), ubiquitination of NPM/B23 could lead to many possible outcomes. Instead of ubiquitin-mediated degradation of NPM/B23, it is possible that ubiquitination changes the confirmation of NPM/B23 and results in NPM/B23 dissociation from the centrosomes. For example, it has been shown that CDC48, a chaperone-like protein, binds to ubiquitinated proteins such as transcription factor
Spt23p90, allowing Spt23p90 to dissociate from its inactive membrane-anchored form and be imported into the nucleus (12,13). Thus, it is possible that upon ubiquitination,
NPM/B23 binds a protein that allows ubiquitinated NPM/B23 to dissociate from the centrosomes, allowing centrosome duplication to proceed. NPM/B23 could possibly then be recruited back to the centrosomes during mitosis through de-ubiquitination.
Ubiquitination has also been shown to regulate the cellular localization of proteins. For example, MDM2 (RING finger ligase) is thought to regulate the nuclear localization of p53 through ubiquitination. Ubiquitination of p53 by MDM2 induces a conformational change in p53 and leads to the unmasking its NES, thus resulting in the nuclear export of p53 (14). Thus, it is possible that once NPM/B23 is ubiquitinated, it
132 undergoes a conformation change unmasking the NLS of NPM/B23, and is imported into the nucleus.
As discussed in chapter IV, the BRCA1-BARD1 complex was shown to ubiquitinate NPM/B23. Our laboratory is currently examining whether this complex is the ubiquitin ligase responsible for the CDK2/Cyclin E phosphorylation-dependent ubiquitination of NPM/B23 observed in this study.
NPM/B23 at the nuclear speckles
In addition to its role in centrosome duplication, we have also demonstrated a role for Thr199-phosphorylated NPM/B23 in pre-mRNA splicing. NPM/B23 is abundantly expressed in the nucleolus, whereas phospho-Thr199 NPM is localized to the nuclear speckles, where splicing factors are present. Moreover, we have found that the phosphorylation of NPM/B23 by CDK2/Cyclin E represses pre-mRNA splicing activity; however, the mechanism of this repression is not yet understood. Several studies have shown hypo- and hyper-phosphorylation of splicing factors results in the repression of their pre-mRNA splicing activity (15). For example, the phosphorylation of the splicing factor ASF/SF2 influences its interaction with other splicing factors and alters pre-mRNA splicing activity (16,17). Another splicing factor, SRrp86, has been shown to inhibit splicing by directly interfering with other splicing factor interactions (18,19). Thus, it is possible that NPM/B23 may act as repressor of splicing by competitively binding to other splicing factors.
Another possible mechanism of splicing repression by phospho-Thr199 NPM is through its RNA binding activity. Pre-mRNA splicing is not only regulated by splicing
133 factors but also by pre-mRNA sequence elements. The exonic or intronic RNA
sequences have been shown to act as splicing enhancers or inhibitors to splicing factors to
modulate pre-mRNA splice site selection (20-22). In Drosophila, P element somatic
inhibitor (PSI) has been shown to interact with a negative regulatory element in the 5’
exon, and this interaction leads to the inhibition of pre-mRNA splicing (23). It has also
been shown that phosphorylation of certain splicing factors, such as SRp40, determines
the RNA-binding specificity of proteins (24). Since phosphorylation of NPM/B23 by
CDK2/Cyclin E enhances its RNA binding activity, it is possible that the phospho-Thr199
NPM may bind specifically to a negative regulatory element in the pre-mRNA and thus repress splicing activity.
NPM/B23 is most likely to be involved in pre-mRNA splicing through its molecular chaperone activity. Proteins with molecular chaperone activity have been shown to localize to the nuclear speckles (25). Components of the spliceosome, such as
Sm-like (Lsm) proteins and La protein, which have been shown to have molecular chaperone-like activity, have been shown to affect pre-mRNA splicing (26-28). Studies have shown these Lsm proteins are required for U6 snRNA (a core component of the spliceosome) (27,29) stability. Recently, Lsm proteins were also shown to support the assembly or remodeling of the spliceosome complex through its chaperone activity (26).
NPM/B23 has also been shown to contain molecular chaperone activity as it can prevent protein aggregation and also promote the renaturation of denatured proteins (30).
Therefore, one could speculate that NPM/B23 acts as a chaperone protein in the pre- mRNA splicing process and perhaps supports the assembly or remodeling of the spliceosome complex. Because NPM/B23 associates with various splicing factors, it
134 could function as a scaffolding protein by bringing different splicing factors together.
Once NPM/B23 is phosphorylated by CDK2/Cyclin E, it loses its association with these splicing factors, perhaps as a result of a conformational change due to the phosphorylation event. This loss of association could result in the disassembly of the spliceosome complex. It has been shown that if the spliceosome is not properly assembled onto the pre-mRNA, pre-mRNA degradation by RNase could occur (Mayeda personal communication). From the in vitro splicing assay, we have observed that the addition of phospho-Thr199 NPM destabilizes the pre-mRNA substrate. Thus, it is possible that NPM/B23 plays a role in the spliceosome assembly.
The involvement of phosphorylated NPM/B23 in splicing brings up an interesting connection between cell cycle progression and pre-mRNA splicing. Since CDK2/Cyclin
E is important for the transition from G1 to S phase of the cell cycle, it would be interesting to determine whether phosphorylated NPM/B23 represses the mRNA splicing of cell cycle proteins. It is possible that phospho-Thr199 NPM represses proteins involved in the G1 phase in order for the cell to progress to S phase. Until now, cell cycle progression was primarily known to be controlled by the degradation of cell cycle related proteins; perhaps NPM/B23 could be acting at the mRNA level as a quick switch between the two phases.
In summary, NPM/B23 has been shown to play a role in many unrelated cellular events. However, the different roles may be explained by one aspect of NPM/B23, its molecular chaperone activity. NPM/B23 may act as a molecular chaperone to prevent aggregation of proteins in crowded environments, such as the nucleolus, centrosomes, and nuclear speckles, and this may explain the diverse localizations and functions of
135 NPM/B23. Proteomic studies based on mass spectrometry have suggested that
approximately 500 proteins are components of the centrosome (31), and over 300
proteins are components of the spliceosome (32). Both the spliceosome and centrosome
are dynamic complexes that constantly undergoes changes in protein-protein interactions
and conformation changes, thus it is not surprising to find NPM/B23 functioning in these complexes.
136 References
1. Tokuyama, Y., Horn, H. F., Kawamura, K., Tarapore, P., and Fukasawa, K.
(2001) J Biol Chem 276, 21529-21537
2. Boveri, T. (ed) (1914) Zur Frage der Entstenhung maligner Tumoren, Waverly
Press, Baltimore
3. Fukasawa, K., Choi, T., Kuriyama, R., Rulong, S., and Vande Woude, G. F.
(1996) Science 271, 1744-1747
4. Fukasawa, K., Wiener, F., Vande Woude, G. F., and Mai, S. (1997) Oncogene 15,
1295-1302
5. Okuda, M., Horn, H. F., Tarapore, P., Tokuyama, Y., Smulian, A. G., Chan, P. K.,
Knudsen, E. S., Hofmann, I. A., Snyder, J. D., Bove, K. E., and Fukasawa, K.
(2000) Cell 103, 127-140
6. Zatsepina, O. V., Rousselet, A., Chan, P. K., Olson, M. O., Jordan, E. G., and
Bornens, M. (1999) J Cell Sci 112 ( Pt 4), 455-466
7. Freed, E., Lacey, K. R., Huie, P., Lyapina, S. A., Deshaies, R. J., Stearns, T., and
Jackson, P. K. (1999) Genes Dev 13, 2242-2257
8. Gstaiger, M., Marti, A., and Krek, W. (1999) Exp Cell Res 247, 554-562
9. Wigley, W. C., Fabunmi, R. P., Lee, M. G., Marino, C. R., Muallem, S.,
DeMartino, G. N., and Thomas, P. J. (1999) J Cell Biol 145, 481-490
10. Schnell, J. D., and Hicke, L. (2003) J Biol Chem 278, 35857-35860
11. Passmore, L. A., and Barford, D. (2004) Biochem J 379, 513-525
12. Rape, M., Hoppe, T., Gorr, I., Kalocay, M., Richly, H., and Jentsch, S. (2001)
Cell 107, 667-677
137 13. Bays, N. W., and Hampton, R. Y. (2002) Curr Biol 12, R366-371
14. Shcherbik, N., and Haines, D. S. (2004) J Cell Biochem 93, 11-19
15. Prasad, J., Colwill, K., Pawson, T., and Manley, J. L. (1999) Mol Cell Biol 19,
6991-7000
16. Xiao, S. H., and Manley, J. L. (1998) Embo J 17, 6359-6367
17. Xiao, S. H., and Manley, J. L. (1997) Genes Dev 11, 334-344
18. Labourier, E., Bourbon, H. M., Gallouzi, I. E., Fostier, M., Allemand, E., and
Tazi, J. (1999) Genes Dev 13, 740-753
19. Barnard, D. C., and Patton, J. G. (2000) Mol Cell Biol 20, 3049-3057
20. Smith, C. W., and Valcarcel, J. (2000) Trends Biochem Sci 25, 381-388
21. Graveley, B. R. (2000) Rna 6, 1197-1211
22. Adams, M. D., Rudner, D. Z., and Rio, D. C. (1996) Curr Opin Cell Biol 8, 331-
339
23. Amarasinghe, A. K., MacDiarmid, R., Adams, M. D., and Rio, D. C. (2001) Rna
7, 1239-1253
24. Tacke, R., Chen, Y., and Manley, J. L. (1997) Proc Natl Acad Sci U S A 94, 1148-
1153
25. van Rijk, A. E., Stege, G. J., Bennink, E. J., May, A., and Bloemendal, H. (2003)
Eur J Cell Biol 82, 361-368
26. Verdone, L., Galardi, S., Page, D., and Beggs, J. D. (2004) Curr Biol 14, 1487-
1491
27. Mayes, A. E., Verdone, L., Legrain, P., and Beggs, J. D. (1999) Embo J 18, 4321-
4331
138 28. Pannone, B. K., Xue, D., and Wolin, S. L. (1998) Embo J 17, 7442-7453
29. Achsel, T., Brahms, H., Kastner, B., Bachi, A., Wilm, M., and Luhrmann, R.
(1999) Embo J 18, 5789-5802
30. Szebeni, A., and Olson, M. O. (1999) Protein Sci 8, 905-912
31. Andersen, J. S., Wilkinson, C. J., Mayor, T., Mortensen, P., Nigg, E. A., and
Mann, M. (2003) Nature 426, 570-574
32. Jurica, M. S., and Moore, M. J. (2003) Mol Cell 12, 5-14
139
Chapter VI
Future Directions
140 The studies elucidating the role of NPM/B23 in the initiation of centrosome
duplication have advanced the centrosome field. Studies to further determine the
function of NPM/B23 at the centrosome are critical. The centrosome consists of
numerous proteins, thus it will be essential to identify NPM/B23 binding partner(s) at the
centrosome. Since NPM/B23 contains different functional domains, the identification of
NPM/B23’s centrosome binding region would give a better understanding on how this
protein functions at the centrosome. Furthermore, it remains unclear how NPM/B23 re-
associates with the centrosome during mitosis. NPM/B23 was shown to be
phosphorylated by CDK1/Cyclin B, a mitotic CDK/Cyclin complex, hence, it is possible
that the phosphorylation by CDK1/Cyclin B on Thr234 and Thr237 is necessary for its
re-association.
Further studies on the ubiquitination of NPM/B23 are also necessary. Because
diverse functions have been attributed to the different ubiquitin moieties, it will be
important to identify the specific type of ubiquitin moiety that is conjugated to NPM/B23.
The identification of the E3 ubiquitin ligase that is responsible for NPM/B23’s
ubiquitination at the centrosome would also further our understanding. Since NPM/B23
is ubiquitinated upon phosphorylation by CDK2/Cyclin E, the ubiquitination of
NPM/B23 might prevent NPM/B23’s re-association with centrosomes during S and G2
phase. Thus, it would be important to identify the ubiquitination region(s) on NPM/B23
and then examine the localization pattern of NPM/B23 at the centrosome.
In regard to NPM/B23’s role in mRNA processing, our finding that phospho-
Thr199 NPM is able to repress pre-mRNA splicing has opened up a new perspective toward cell cycle progression and mRNA processing. It would be interesting to examine
141 whether phospho-Thr199 NPM represses the splicing reaction itself or rather if it inhibits the assembly of spliceosome complex. Also, as discussed in Chapter V, RNA binding could affect NPM/B23’s function in pre-mRNA splicing. Thus, the identification of the
RNA-binding sequence of NPM/B23 could further our understanding on how NPM/B23 is repressing pre-mRNA splicing. Moreover, it will be essential to identify the target genes repressed during mRNA processing by NPM/B23. Gene expression microarrays should be performed to analyze the relative quantities of cell cycle-related mRNAs in cells expressing NPM/wt, NPM/T199A or NPM/T199D.
142