Citation
Differential activities of cyclin-dependent kinases in the phosphorylation of the C-terminus of p70 S6 Kinase

Material Information

Title:
Differential activities of cyclin-dependent kinases in the phosphorylation of the C-terminus of p70 S6 Kinase
Creator:
Papst, Phillip J
Publication Date:
Language:
English
Physical Description:
xii, 97 leaves : illustrations ; 29 cm

Subjects

Subjects / Keywords:
Protein kinases ( lcsh )
Phosphorylation ( lcsh )
Phosphorylation ( fast )
Protein kinases ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 90-97).
General Note:
Submitted in partial fulfillment of the requirements for the degree, Master of Arts, Biology.
Statement of Responsibility:
by Philip J. Papst.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
37886912 ( OCLC )
ocm37886912
Classification:
LD1190.L45 1997m .P37 ( lcc )

Downloads

This item has the following downloads:


Full Text
DIFFERENTIAL ACTIVITIES OF CYCLIN-DEPENDENT KINASES
IN THE PHOSPHORYLATION OF THE C-TERMINUS
OF P70 S6 KINASE,
by
Philip J. Papst
B.A., University of Colorado, 1989
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Master of Arts
Biology
1997


This thesis for the Master of Arts
degree by
Philip Jules Papst
has been approved
by
Naohiro Terada
Linda Dixon
?//<£/? 7
Date
Martin Klotz


Papst, Philip, J. (MA., Biology)
Differential Activities of Cyclin-Dependent Kinases in the
Phosphorylation of the C-Terminus of P70 S6 Kinase.
Thesis directed by Assistant Professor Naohiro Terada
Abstract
p70 S6 Kinase is the major physiological kinase that
phosphorylates S6 protein of the 40 S subunit of eukaryotic
ribosomes. p70 S6 kinase (p70S6K) has a set of phosphorylation
sites located in the carboxy-terminus at positions Ser 411, 418,
424 and Thr 421 that are phosphorylated by an unknown
mechanism. These sites represent a recognition motif found in
substrates of cyclin-dependent kinases (CDKs). It had been noted
previously that cell cycle regulatory protein GDK1 can
phosphorylate the carboxy-terminal sites in vitro. Here I
investigated whether this activity is restricted to CDK1 or more
general to other cell cycle regulatory proteins like CDK2 since
both CDK1 and CDK2 share a high level of amino acid similiarity
and similiar phosphorylation activity toward Histone H1. Using
recombinant baculovirus technology, I have generated wild-type
and mutant p70 S6 Kinase enzymes that provide suitable
subtrates to investigate differential phosphorylation events. I
m


demonstrate that a Xenopus CDK1/cyclin B complex but not a
CDK2/cyclin E complex phosphorylates all the sites at the C-
terminus of p70 S6 kinase, especially at serine 411 in the
KIRSPRR sequence. The CDK1/cyclin B complex phosphorylated
the KIRSPRR peptide with an apparent Km of 50 jiM; whereas
CDK2/cyclin E complex did not phosphorylate the peptide well
(Km > 500 p,M). Additionally, CDK1/cylin B complexes
immunoprecipitated from the lysates of proliferating human T
lymphocytes were able to phosphorylate the KIRSPRR peptide very
well, however, immunoprecipitated CDK2, 4, or 6 complexes
showed poor phosphorylation activity toward the peptide.
p13suc-immune complexes which associate with both CDK2 and
CDK1 phosphorylate histone H1 peptide (KKSPKK) at about the
time when CDK2 is activated in cell cycle progression of primary
T lymphocytes (the G1/S phase of the cell cycle). By contrast
phosphorylation of the KIRSPRR peptide of p70 S6 kinase is
increased later in parallel to CDK1 activation, at the G2/M
boundary. Taken together, these data demonstrate differential
specificity of CDK1 and CDK2 on a specific peptide sequence of
p70S6K, and support the idea that CDK1 may play a role in the
regulation of p70S6K.
IV


This abstract accurately represents the content of the
candidate's thesis. I recommend its publication.
Signed
Naohiro Terada
V


DEDICATION
I would like to dedicate this thesis to my family.
I wish to dedicate this thesis to my wife, MaryLyn. I wish to
thank her for her love and continual support. I will be forever
grateful for her kind words of encouragement and inspiration. My
education has taken up a large part of our lives, and thanks
MaryLyn for being there every step of the way. Thanks also to my
children, Colton and Mallory, for their love and joy they show me
every day.
I wish to thank my parents Linda and Junior Manning for
their never ending support and especially for the financial
support in my last couple semesters. I could not have completed
the degree program without all of your help, love, and
encourgement.
I also would like to thank my father and mother-in-law,
Carolyn and Wes Wilklow for all of their general support and
especially for their financial assistance towards the end of my
degree program. Thank you for always being there when I needed
someone to talk to.
Finally, I would like to thank Kelly and Holly Miller for the
donation of their computer. They made it possible for me to
spend countless hours working with this computer at home, and
since 1 was not in a position to purchase a computer at the time,
their thoughtfulness made a huge difference during the writing of
this manuscript. Thank you so very much, I could not have done it
without your help.


ACKNOWLEDGMENTS
I would like to thank Dr. Naohiro Terada for all his support
and guidance with my thesis project these past few years. He has
taught me the skills and techniques to be a good scientist. I am
forever grateful for your patience and friendship you have shown
me throughout the pursuit of my Master's degree. One of the best
ways to expand your intellect, is to learn by example. Dr. Terada
has shown me the best example of an impressive, extremely
bright, and hard working scientist who knows how to get the job
done. At the same time exhibiting the compassion and
understanding needed in the training of a graduate student. Thank
you, Dr. Terada, for the opportunity to work for you and thank you
for being such a great mentor.
I would like to thank Dr. Linda Dixon for her help with
University guidelines and policies. She has been an excellent
academic advisor and has guided me through my long and
challenging Master's degree program with patience and
compassion. Thank you Dr. Dixon for going that extra mile for me
and keeping me on track with all the requirements necessary to
complete this endeavor.
I would also like to thank Dr. Martin Klotz for his critical
analysis and helpful commentary of my thesis manuscript. I am
also very grateful for his critique of my thesis defense
presentation and organization. Thanks to your helpful comments,
I have learned an enormous lesson about scientific
communication.
I would also like to acknowledge Dr. Masayuki Nagasawa and
Dr. Hirotaka Sugiyama for their technical help and assistance in
the laboratory. Dr. Sugiyama helped with the initial construction
and cloning of the wild-type and mutant p70 S6 Kinase proteins.


Dr. Nagasawa provided assistance for many of the experiments
presented here and especially with the phosphorylation assays.
Finally, I would like to thank Jacque Horvath for her
computer knowledge and technical writing assistance during
preparation of the manuscript. Thanks to her I learned a fair
amount about Microsoft Word and all of its capabilities. Thank
you Jacque for always taking the time to answer all my questions.


CONTENTS
CHAPTER
1. INTRODUCTION........................................ 1
Background............................................ 1
Protein Serine/Threonine Kinases...................... 2
The Cell Division Cycle............................... 3
The History of p70 S6 Kinase.......................... 7
p70 S6 Kinase and Translation.................... 1 1
The Use of Rapamycin as a Tool to Investigate
the Role of p70 S6 Kinase in Cell Proliferation... 1 4
Signaling Pathways.................................... 1 5
The p70 S6 Kinase Molecule: the Domains of
the Kinase and Their Putative Functions............... 1 7
Model of p70,S6 Kinase Regulation..................... 2 0
The CDK Substrate Recognition Domain.................. 2 3
The Discrepancy....................................... 2 4
Hypothesis............................................ 2 4
ix


2. MATERIALS AND METHODS.................................... 2 6
Cell Culture.......................................... 2 6
Antibodies and Immunoprecipitations................... 2 7
Baculovirus Expression System......................... 2 8
Construction of the p70 S6 Kinase Clones.............. 2 9
Transformation and Transposition of BAC p70
Clones into BAC 10 Cells for Integration of p70
S6 Kinase Into the Baculovirus Genome.... 3 1
Isolation and Analysis of Recombinant p70 S6
Kinase Bacmid DNA for Transfection and Viral
Production............................................ 3 2
Transfections....................................... 3 3
Infections............................................ 3 4
CDK1 /Cyclin B Phosphorylation of
Recombinant p70 S6 Kinase............................. 3 5
Phosphopeptide Mapping................................ 3 6
Protein Precipitation and Digestion............. 3 6
Electrophoresis/Chromatography.................. 3 7
Synthetic p70 S6 Kinase Ser 411
Phospho-Peptide Map............................. 3 8
CDK/Cyclin In Vitro Kinase Assay...................... 3 9
x


T Cell Time Course............................. 4 0
T Cell CT-p70 S6 Kinase In Vivo
Phosphorylation................................ 4 0
Activity Time Course of p70 S6 Kinase
andCDKl........................................ 4 1
Cell Cycle FACS Analysis....................... 4 2
Increase of CT-p70 S6 Kinase Phosphorylation
by Nocordizole Treatment In Vivo..................... 4 2
3. RESULTS................................................. 4 4
Baculovirus Expression System Provides High
Quantities of Functionally Active Wild-Type or
Mutant p70 S6 Kinase for Phosphorylation
Analysis............................................. 4 5
CDKl/Cyclin B and Not CDK2/Cyclin E
Phosphorylates Recombinant CT-p70 S6
Kinase on Ser 411.................................... 4 8
The Rate of Peptide Phosphorylation by
CDKl/Cyclin B Complex is Higher Than
CDK2/Cyclin E Complex................................ 5 7
CDKl/Cyclin B Preferentially Phosphorylates
CT-p70 S6 Kinase Whereas Other Members of
the Cyclin- Dependent Kinase Family Do Not..... 6 1
Differential CDK 1 Activity on CT-p70 S6
Kinase C-Terminal Peptide at a Different Cell
Cycle Phase Than Histone HI Peptide.................. 6 3
xi


Late Cell Cycle Phosphorylation Versus Early
Activity of p70 S6 Kinase........................... 6 5
The Enhancement of CDK1 Protein Expression
and Activity With Nocordizole Treatment
Also Markedly Increases p70 S6 Kinase
Phosphorylation In Vivo............................. 7 2
4. DISCUSSION............................................. 7 6
Implications for CDK1 Phosphorylating p70 S6
Kinase.............................................. 7 7
Subcellular Localization............................ 7 8
Nucleolar Organization.............................. 7 9
Future Experiments.................................. 8 0
APPENDIX.................................................. 8 2
Equipment........................................... 8 2
Chemicals........................................... 8 4
Cells............................................... 8 7
Reagents............................................ 8 7
Solutions........................................... 8 8
REFERENCES................................................ 9 0
xii


CHAPTER 1
Introduction
Background
One of the most fundamentally important aspects of biology
is how cells grow and proliferate and how signals generated at
the plasma membrane by various stimuli are transmitted to the
nucleus, whereby the appropriate response can be made. The
events leading up to and encompassing cell division are
enormously complex (Patel et.al. 1991). To understand how cells
communicate, it is essential to first isolate individual molecular
interactions that make up a particular response and then analyze
it in the context of other reactions occuring as a result of the
stimulus. Cell surface receptor ligation induces a cascade of
phosphorylations and dephosphorylations transduced by kinases
and phosphatases. Many of these enzymes function in series as
part of a specific pathway resulting in increased transcriptional
and translational activity required for proliferation. Several of
these pathways function in parallel generating a complex network
of signal transductions all induced by the same growth stimulus.
l


One of the goals of molecular biology today is to dissect these
pathways to understand how cells communicate at the molecular
level. The molecular enzymology of the protein kinases mediating
these communications will provide the necessary framework upon
which to acheive these goals.
Protein Serine-Threonine Kinases
There are two major eukaryotic families of protein kinases,
the tyrosine kinases and the serine/threonine kinases (Hanks et
al. 1988). These kinases have the ability to add phosphorous to
specific tyrosine, serine or threonine amino acids of other
proteins thereby activating them to carry out their specific
functions. The activity of serine-threonine kinases play an
important role in eukaryotic cell physiology including control of
cell proliferation. Unlike the tyrosine kinases which are more
implicated with phosphorylation events coupled to cell surface
receptor, serine/threonine kinases function intracellularly,
phosphorylating numerous proteins in series that function as
downstream effectors of signal transduction pathways. There are
several members of this family including protein kinase C, the
cAMP-dependent protein kinases, the cyclin-dependent kinases
controlling the cell cycle, and the Raf signalling complex. Raf is
an excellent example of a serine-threonine kinase that is
activated downstream of the G protein Ras and then serving to
2


activate a protein kinase cascade involving the MAP kinase
pathway that ultimately triggers cell proliferation. Protein
serine-threonine kinases are more abundant in the cell than their
tyrosine kinase counterparts. Amoung these serine-threonine
kinases is the enzyme called p70 S6 Kinase (p70S6K), which lies in
a signal transduction pathway distinct from the MAP kinases
(Ming et al. 1994). As will be discussed in more detail later,
p70 is a key mediator in the regulation of ribosome biogenesis
and the subsequent regulation of protein synthesis that is
required for cell proliferation.
The Cell Division Cycle
Cell growth is divided into distinct phases, each
representing different functions as the cell prepares to go
through mitosis. The cell cycle is made up of G1 (gap 1), a
preparatory phase before S phase (DNA sythesis), and G2 (gap 2),
which prepares the cell for mitosis (M). Growth factors stimulate
cells to progress through this cycle, which is controlled by a
family of protein serine-threonine kinases that are highly
conserved through evolution. These kinases are called cyclin
dependent kinases (CDKs) because they associate with proteins
referred to as cyclins (Figure 1). CDKs are thought of as the
active component of the CDK/cyclin complex. Together, they
along with a host of other regulatory molecules help guide the
3


entry into, progression of, and exit from the cell division cycle
(Pines J. 1993). More importantly, they function at two main
controlling points, of the cycle. They control the progression of
cells from G1 to S and from G2 to M. When quiesent cells receive
a signal from a growth factor or other mitogen, different distinct
combinations of CDKs and their associated cyclins assemble into
complexes that provide the necessary kinase activity to traverse
through the control points and progress through the cell cycle.
4


Figure 1. Differential roles of cyclin-dependent kinases and their
associated cyclins in cell cycle regulation.
5


The two main CDK's studied in this project, are CDK1 (also
referred to as p34cdc2) and p33CDK2. CDK1 associates with
either cyclin A or cyclin B and is active in G2 and M phases of the
cycle. The G1/S transition is controlled by CDK2 and its cyclins,
either cyclin A or cyclin E (Elledge et al. 1991). The initiation of
the growth cycle is regulated by the D type cyclins and its
association with CDK4 and CDK6 (Pine J. 1993; Sherr C.J. 1993).
In each case, the components of the complex physically interact
with each other, subsequently becoming active specifically
phosphorylating various substrates that coordinate numerous
transcriptional activities that prepares the cell for division.
Then the complex dissociates in time for the next family of
CDK/cyclin complexes to carry out their function. Substrate
specificity is an important attribute of each CDK/cyclin complex.
CDK1 is responsible for the breakdown of the cytoskeleton and
the nuclear envelope by phosphorylating many of their structural
components. CDK1 also phosphorylates Histone H1 thereby
regulating chromatin structural components in M phase (Migg E.
1993). Alternatively, CDK2 and its associated cyclins
phosphorylate a protein complex called Rb ( the retinoblastoma
protein), which is known to be an inhibitor of the transcription
factor E2F. The phosphorylation of Rb releases E2F and allows it
to bind to specific promoter regions of DNA in the nucleus thereby
initiating new gene expression needed for cell division (Nigg E.
6


1993). Given the importance of CDK1 and CDK2 in regulating the
G2/M and G1/S transitions, the search for other substrates of
these serine/threonine kinases might lead to new pathways of
regulation important for cell proliferation.
The History of p70 S6 Kinase
In the late 1980's, due to the role of protein
phosphorylation cascades involved in the activation of cell
growth and the importance of ribosomal S6 protein for the
regulation of the mRNA translation and the control of protein
sysnthesis, it was imperative to identify the kinases controling
these processes. A protein was found in cell extracts with potent
S6 activity that itself was sensitive to phosphatase treatment
indicating that this protein itself was phosphorylated (Jeno et al.
1988). The putative S6 activity was isolated, purified to
homogeneity, and identified as a kinase with the apparent
molecular weight of 70 kilodaltons. Upon cloning, two
transcripts were identified, differing only in 23 amino acids at
the amino terminus (Grove et al. 1991). This other transcript
produced by differential mRNA splicing, was called p85S6K (
Banerjee et al. 1990). It contains a cluster of basic amino acids
(Arg-Arg-Arg-Arg-Arg-Arg) that comprises a nuclear
translocation motif in these 23 amino acids. Nuclear localization
of p85S6K was shown by using microinjected expression vectors
7


S6K
containing p85 or fusion proteins containing the first 150
SgK
amino acids of p85 (Reinhard et al. 1994). It has been
estimated that approximately 4000-6000 p70 molecules are
expressed in a single cell, as compared to only about 300
molecules of p85 (Kozma et al., 1993). It was proposed that
this form of the enzyme maintains a pool of phosphorylated S6
protein in the nucleus but its actual function is unknown (Franco
and Rosenfield, 1990). The p70 constituent is active primarily in
the cytoplasm and is now known to be the major physiological
kinase that phosphorylates S6 in the cleft of the assembled
ribosome where the mRNA is bound and protein synthesis is
initiated (Ballou et al. 1991, Blenis et al. 1991).
There are actually two families of S6 kinases: the p90rsk
encoded kinases and the p70 and p85 S6 kinases. The p90rsk
family of kinases only phosphorylate S6 kinase in vitro. The
p70/85 S6 kinases are the major physiological kinases that
phosphorylate S6 in vitro and in vivo. There is considerable
evidence that places the two different kinases on distinct
signalling pathways. Unlike p70S6K, p90rsk is phosphorylated and
activated by the p42 and p44 mitogen-activated protein kinases
(MAPK) in response to signals transmitted from the
RAS/RAF/MEKK series of kinases that are activated by growth
factor receptor stimulation (Ming et al. 1994; Blenis et al. 1991).
8


The p90rsk and MAP Kinases then translocate to the nucleus to
phosphorylate transcription factors, thereby promoting
transcription of genes required for proliferation. The p70 S6
kinases are completely distinct from this signalling pathway and
have a function all their own. The p70S6K is known to be in a
signalling pathway that places it below the lipid kinase PI3K
(Figure 2).
9


Figure 2. Overview of growth factor mediated signal
transduction pathways leading to gene expression or protein
synthesis. p70 S6 Kinase lies in a distinct pathway from the
MAP kinases.
>
10


p70S6 Kinase is a serine/threonine intracellular kinase that
is inactivated in quiescent cells, and upon activation has been
shown to phosphorylate the ribosomal S6 protein in higher
eukaryotes (Ballou et al. 1991; Blenis et al. 1991). This enzyme
is an intermediary in a signal transduction cascade of
phosphorylations initiated at the plasma membrane by many
known mitogenic stimuli including EGF, PDGF, IL-2, IL-3, GM-CSF,
and insulin. This kinase has been shown to be important for
quiescent cells to enter the cell division cycle (G0/G1), as
indicated by the inhibition of p70 by the specific inhibitor
rapamycin or a micro-injected antibody against p70 (Lane et al.
1993; Chung et al. 1992). These inhibitors either arrest the cell
in G1 or cause G1 prolongation when cells have already entered
the cell cycle (Terada et al. 1993). P70S6 Kinase specifically
phosphorylates multiple sites on ribosomal S6, a protein of the
40S ribosomal subunit, and thus is coordinately involved in the
regulation of protein synthesis (Bandi et al. 1993).
P70 S6 Kinase and Translation
S6K
p70 phosphorylates ribosomal protein S6 (Ballou et al.,
1991, Blenis et al., 1991). The ribosomal S6 protein is located in
a position in the 40 S subunit that is proposed to be the major
functional site involving mRNA and tRNA binding (Bommer et al.,
11


al.,
be
an
1930; Terao and Ogata,1979; Tolan and Traut, 1981; Geyer et al.
1932; Kjeldgaard et al. 1994). After mitogenic stimulation,
p7pS6K phosphorylates several sites in the carboxy-terminus of
S6 a region proposed to lie in the cleft where mRNA binds,
suggesting a role in the regulation of translation (Kozma et
1989). Nielson et al. suggested that S6 phosphorylation may
involved in the recruitment of mRNA's into polysomes because
ibodies to S6 block the binding of elf-2/GTP/MET-tRNA
complexes to 40 S ribosomes. This suggested that phosphorylated
S6 may be directly involved in translation initiation and directly
modulate mRNA binding in ribosomes during protein systhesis.
Th us increasing p70S6K activity increases S6 phosphorylations,
subsequently increases translational activity, especially peptide-
chain initiation.
The correlation to translational regulation is not associated
witi overall protein synthesis, because there is a discrepancy
between S6 phosphorylation and the overall rate of protein
syrithesis( Thomas et al., 1980,1982). Many cell lines like Swiss
3T3 or Cos cells when stimulated with EGF or other growth
factors increase S6 phosphorylation rapidly (within 30-60
minutes), and then as cell cycle progresses the phosphorylation of
S6 gradually decreases. Overall protein systhesis however,
increases more slowly reaching a maximal point after 3-4 hours.
Therefore, it was suggested that S6 phosphorylation is
functioning in a mechanism of selective translational control
12


instead of regulating overall protein synthesis( Thomas et al.
1992). Subsequently, this established the current dogma that
S6K
p70 regulates translation of specific mRNA's through the action
of S6 phosphorylation. Taking this notion to a higher level, a
mechanism was proposed for this kind of regulation. All
ribosomal proteins in higher eukaryotes have a consensus
sequence at their transcription initiation site (Jefferies et al.
1994). This sequence typically consists of a span of pyrimidines
that follows a cysteine in the transcription initiation site
designated CYYYYYY. This sequence is followed by a GC rich
region. Interestingly, in mammalian cells, both elongation factor
1a (eEF-1a) and elongation factor 2 (eEF-2) have this same
consensus sequence at the 5'-terminus of their mRNA, but not in
their genomes (Nakanishi et al., 1988; Uetsuki et al., 1989). It is
possible then that this sequence is correlated to selective
translational regulation and indeed, several groups have since
shown that this 5' polypyrimidine tract is required for the
translational control of r-bosomal protein mRNA's and elongation
factor mRNA's (Levy et al.). Thus the activation of p70S6K leads to
the phosphorylation of S6 which inturn upregulates the
translation of selective mRNA's encoding the components of the
ribosome. The ribosomal proteins and elongation factors need to
be present before general protein systhesis can begin. This then
might possibly explain the discrepancy between early and rapid
13


S6 phosphorylation and later increases in overall protein
systhesis.
The Use of Rapamvcin as a Tool to Investigate the Role of p70 S6
Kinase in Cell Proliferation
Much of the knowledge obtained to date about the regulation
of p70S6K has been deduced from experiments using the antifungal
antibiotic Rapamycin (RAP). Rapamycin is a macrolide with
functional similarities to cyclosporin and FK506; they are all
potent immunosuppresants and have been shown to prolong
survival of organ allografts and prevent autoimmune diseases
(Morris R. E.). RAP inhibits proliferation of lymphocytes as well
as other types of cells by arresting cell cycle progression at
early G1 phase or delaying entry into S phase if mitogen
stimulated quiesent cells have already entered the cell cycle
(Terada et al., 1993). RAP binds to FKBP12 (FK506 binding
protein) and this associates with a third molecule called FRAP or
RAFT, and it is this complex that is responsible for the indirect
S6K
inhibition of p70 ( Brown et al., 1994,1995; Weng et al., 1995;
Schreiber et al. 1991). This FRAP protein has considerable
homology with the catalytic subunit of PI-3 kinase, and has since
been demonstrated by Brown et al to be located upstream of
p70S6K. Our lab and others have demonstrated that the RAP/FRAP
complex specifically dephosphorylates and inhibits the activity
14


S6K
of p70 at Thr 229, Thr 389, and Ser404, with no alterations in
many other early signal transduction events after mitogen
stimulation ( Chung et al., 1992; Kuo et al., 1992; Price et al.,
1992; Calvo et al., 1992; Terada et al., 1992). Indeed, through the
use of recombinant DNA techniques the substitution of Thr 229 to
alanine eliminates the sensitivity of p70S6K to RAP and makes an
inactive kinase (Sugiyama et al., 1996). The Thr 389
phosphorylation site has also been shown to be a Rapamycin
sensitive site (Thomas et al., 1995). Indeed, the motif
surrounding the Thr 389 is highly conserved in many members of
the second messenger family of protein kinases including the
PKC isoforms. This domain has been implicated as a crucial
sequence in the maintenance of kinase activity in these other
kinases. Rap has no effect on the other phosphorylation sites in
the carboxy-terminal region ( Ferrari et al. 1993; Pearson et al.
1995 ).
Signalling Pathways
S6K
The current signalling pathway of p70 and other
associated pathways form a complex network of signal
transduction events during mitogenic or growth factor
stimulation. One of the signalling pathways activated after cell
15


S6K
surface receptors are ligated includes the p70 pathway. The
upstream kinases or phosphatases remain to be clearly identified,
however there is preliminary evidence indicating phosphatidyl
inositol-3 kinase (PI-3 kinase) and protein kinase B are located
upstream of p70S6K (Bos J.L.,1992 and Boudewijn et al., 1992;
Burgering et al. 1995). PI-3 Kinase is one of many lipid kinases
that are part of a pool of activated second messenger molecules
stimulated after EGF and PDGF receptor ligation. PI-3 kinase
phosphorylates phosphoinositol 3,4,5-triphosphate which then
through a series of proposed intermediaries, leads to p70S6K
activation (Downward. J., 1995 ). Neither of these kinases
directly phosphorylate p70S6K either, demonstating other as yet
unidentified kinases involved in phosphorylating. p70S6K. Other
mitogen-stimulated signalling pathways bifurcates at the level
of the p21 RAS, thus making the ERK-encoded MAP kinase 1 and 2,
the p38 MAP kinases, the p42 jun kinase, and pathways
transmitted from protein kinase C (PKC) all distinct signalling
cascades from the p70 signalling pathway (Ming et al., 1994).
16


The p70 S6 Kinase Molecule: the Domains of the Kinase and Their
Putative Functions
S6K
The p70 protein is 502 amino acids long and consists of
several domains important for its function (Banerjee et al.,
1990). These domains were characterized with the help of
rapamycin studies. The cataylic domain comprising residues 89-
355 contains the threonine 229 and Thr 389 sites that are the
principal phosphorylated sites responsible for kinase activity;
the loss of these sites renders the kinase inactive (Sugiyama et
al., 1996; Pearson et al., 1995). The catalytic domain also
exhibits an ATP-binding site (GKGGYG) located at amino acid 98-
103 functionally similiar to other known kinases. It has been
postulated that there are two independent pathways which
regulate kinase activity, group 1, located at positions Thr 229,
Thr 389, and Ser 404 is responsible for its activation, and the
putative translational regulation of specific ribosomal protein
and elongation factor mRNA's (Han et al., 1995; Pearson et al.,
1995) . This phosphorylation is inhibited by the
immunosuppressive drug Rapamycin, but the T229 and T389
phosphorylation sites are differentially regulated by different
kinase kinases that are Rapamycin insensitive (Dennis et al.
1996) . Group 2 is located in the carboxy-terminus at positions
Ser 411,418,424 and Thr 421 (Ferrari et al., 1993; Weng et al.,
1995). This site is rapamycin insensitive. Numerous reports
17


have focused on group 1 phosphorylation site and its involvement
in translational regulation, however, little is known about group
2 phosphorylation activity and its physiological consequences.
Phosphorylation of group 1 occurs early in the cell cycle whereas
the C-terminal site phosphorylation occurs in later stages of the
cell cycle ( Lane et al. 1993; Susa et al. 1989; Weng et al. 1995).
18


Figure 3. Protein structure
rapamycin, wortmannin-sensitive
PI3K-sensitive
19


Model of p70 S6 Kinase Regulation
The culmination of many years of investigation into the
regulation of this kinase has produced a model of that attempts
to explain many of the findings(Banerjee et al. 1993). Residing in
the catalytic domain is the Thr 229 phosphorylation site that is
essential for kinase activity and also contains the ATP binding
site. This site is proposed to be blocked by a portion of the
carboxy-terminus that contains a putative pseudosubtate
sequence. This basic pseudosubstrate site is envisioned to bind
to and occlude the substrate binding site including the ATP-
binding domain and maintain the kinase in a basal-inhibited
state. This carboxy-terminal region is refered to as an auto-
inhibitory domain. There is an acidic region (Asp29-Glu46)
slightly amino terminal of the catalytic domain that is thought to
be a candidate for the complementary substrate binding site; a
region that could provide a suitable binding site for the highly
basic protein ribosomal S6. Thus the basic pseudosubstrate
region in the carboxy-terminal tail is thought to fold over and
inhibit the cataytic domain by binding to this upstream acid
region when the kinase is in the inactive form. Upon activation of
quiescent cells, p70 phosphorylations in and around the C-
terminal pseudosubstate region weakens the electrostatic binding
of the two complementary regions thereby releasing the occluded
20


catalytic domain, permiting access to ATP and S6 and allowing
the enzyme to be activated via the group 1 phosphorylation sites.
The Thr 389 site of group one actually lies in the linker region
and its phosphorylation would also help in this conformational
change that exposes the catalytic site (Pearson et al., 1995).
21


R70s6k
Thr389
Figure 4. Proposed model for the regulation of p70
S6K
22


The CDK Substate Recognition Domain
Each of the four carboxy-terminal phosphorylation sites
comprise the sequence Ser/Thr-Pro motif which is often
recognized as a substrate for CDk-like and MAPK -like proteins (
Banerjee et al., 1990; Kitagawa et al. 1996). The specific
sequence RSPRR of the ser 411 region is the actual consensus
sequence of CDK1 substrate recognition motif ( Songyang et al.,
1994). I speculated that this sequence present in the carboxy-
terminus of p70 had some importance in the regulation of the
enzyme, and thus further explorations into the functional
significance that this might imply might lead to novel mlecular
interactions crucial to the control of proliferation. Given the
evidence to date demonstrating two distinct activation domains,
there must be other kinases active in the cell that phosphorylate
p70 and regulate its activity for other important processes in
cell physiology. Scanning the literature on p70S6K produced one
piece of evidence that CDK1 can indeed phosphorylate these C-
terminal sequences in vitro ( Mukhopahdya et al., 1992). This
paper only showed CDK1 and did not investigate other cell cycle
regulatory proteins. Here I investigate the differential
specificities of the major players of the cell cycle regulatory
proteins in the regulation of p70S6K.
23


The Discrepancy Between the Timing of the Cell Cycle and the
Activation of p70 S6 Kinase
There was an another intriguing observation made in our
laboratory concerning the time course of phosphorylation of p70
in vivo. Many experiments have shown massive phosphorylation of
p70 late in the cell division cycle of lymphocytes. In contrast,
p70 activity in these cells has been shown to be very rapid
after the cells were stimulated to enter the cycle. If the
principal function of the kinase occurs very early and is
associated with G1 progression as the previous studies have
indicated, why does this enzyme need to be so intensively
phosphorylated late in the cell division cycle? I propose that the
answer to this question may be found closely correlated with the
activity of CDK1 and the control of the G2/M transition.
Hypothesis
The purpose of this work is to investigate the carboxy-
terminal phosphorylations of p70 S6 kinase and its potential role
in late cell cycle regulation through the interaction of cyclin-
24


dependent kinases and there associated cyclins. This project
presents and supports the hypothesis that CDK1/cyclin B can
phosphorylate a specific serine residue in the carboxy terminus
of p70 S6 kinase only late in G2/M phase of the cell division
cycle. CDK 1 is the principal CDK invovled in this regulation. Both
in vitro and in vivo evidence is presented to support this
S6K
hypothesis and illustrate a direct link of the p70 signalling
pathway and the cell cycle machinery.
25


CHAPTER 2
MATERIALS and METHODS
Cell Culture
Human peripheral blood cells were obtained by
leukopheresis of healthy donors from Children's Hospital Blood
donor center. Mononuclear cell suspensions were prepared by
Ficoll-Hypaque gradient centrifugation and T cells were obtained
by E- rosette enrichment using nurimidase treated sheep red
blood cells (Colorado Serum Co, Denver, CO) as described. Cells
were maintained in RPMI 1640 complete medium (Gibco BRL,
Grand Island, NY) supplemented with 10% (v/v) heat inactivitated
fetal bovine serum (FBS, Hyclone, Logan, Utah), 2mM Glutamine
(Gibco), 100u/ml penicillin and 100mg/ml streptomycin (Gibco).
T cells were stimulated with Phorbol 12,13-dibutyrate (PDBu,
Sigma, St. Louis,MO) and calcium ionophore, ionomycin
(Calbiochem, San Diego, CA) dissolved in dimethyl sulfoxide
(DMSO). Rapamycin was obtained from the Drug Synthesis and
Chemistry Branch, NCI, and dissolved in PBS to give a 0.1mg/ml
stock solution.
26


Insect cells (Spodoptera Frugiperda Sf9) were maintained at
30 degrees in serum-free SF900 II SFM medium (GIBCO)
supplemented with 50u/ml penicillin and 50mg/ml streptomycin.
Cells are cultured in a monolayer and become suspended in
medium by gentle dispersion. Cells were passed every four days
by 1/3 or 1/4 dilution.
Antibodies and Immunoprecipitations
Antibodies used in this study were; p70 a rabbit
polyclonal antibody raised against the carboxy-terminus (amino-
acids 502-525 NSGPYKKQAFPMISKRPEHLRMNL) anti-CDK1, rabbit
polyclonal antibodie raised against human CDK1 (kindly provided
by David Beach); anti-cdk2, rabbit polyclonal antibodies raised
against an epitope 283-298 of human cdk2 (sc-163, Santa Cruz
Biotechnology, Santa Cruz, CA); anti-CDK4, rabbit polyclonal
raised against amino acids 1-303 (Santa Cruz); anti-Cdk6, rabbit
polyclonal raised against amino acids 306-326 (Santa Cruz);
anti-cyclin A, rabbit polyclonal antibodies raised against human
cyclin A (kindly provided by Jonathon Pines); anti-cyclin B, mouse
monoclonal lgG1 recognizing epitope 1-21 of human cyclin B1
(GNS-1 ,Pharmingen, San Diego, CA). Immunoprecipitations were
done using Protein G sepharose (Zymed Laboratories, South San
Francisco, CA)and p13 sue agarose beads (Oncogene Science,
27


Uniondale, NY) Typically 5-10 pi of antibody was used per 0.5-
1.0ml of lysed cells in lysis buffer ( RIPA buffer or P70 S6K buffer)
for 1-2 hours on ice. Prewashed protein G sepharose (30-50pl)
was added and rotated at 4 C for 30 minutes-1 hour. Separose
beads were centrifuged and washed twice in lysis and then 1x
loading dye was added and the samples boiled for 5 minutes to
remove proteins from sepharose.
Baculovirus Expression System
Baculovirus expression system was employed to generate
large quanties of active and inactive wild type and mutant protein
which provided excellent substrates for peptide phosphorylation
analysis. This is a viral expression method in armyworm larval
insect cells. The insect cells are infected with recombinant
Autogragha California nuclear polyhedrosis virus (AcNPV) by
means of efficient vectors that allow high levels of heterologous
gene expression. Baculoviruses have a restricted host range, only
infecting specific invertebrate species. Large quanties of soluble
functionally active protein of interest that can be selected
specifically by a 6-histidine tag sequence located 5' of the cloned
mammalian protein. The vectors employed also have a powerful
polyhedron promoter, transposon recognition sequences, a rTEV
28


protease recognition sequence, a large multi-cloning site, and
selectable antibiotic markers. Recombinant DNA clones were
isolated and used to transform Bac 10 cells that contains the
viral genome. These cells contain the 150kb viral genome and
helper plasmid that encodes the transposase enzyme. The DNA
between the tn7 sequences including the p70 S6K sequence will
transpose into the viral genome at a specfic target that is under
the control of lac Z. This allows the inducible expression by
IPTG and blue gal and the selection of white recombinant clones
against a background of non-recombinant blue colonies. White
clones are restreaked to ensure purity and the DNA isolated from
these possible positive clones. Large molecular weight DNA is
analyzed by 0.5% agarose gel. This DNA is then used for the
transfection of SF9 insect cells to obtain complete virus
particles that are then used to infect SF9 cells on a large scale.
Construction of P70 S6 Kinase Clones
S6K
The full length p70 cDNA clone was kindly provided by
George Thomas and initially subcloned into pXST mammalian
expression vector. This pXST clone was analyzed by restiction
endonuclease cleavage and mammalian transfection to determine
efficacy of the clone (Sugiyama et al.,1996). The Sal I restriction
endonuclease cleaved a 1.6 kb p70S6K fragment from these pXST
clones that represents the wild type used for further subcloning.
29


I purified the fragment by separation through an agarose gel with
subsequent DNA purification with the Quiagen purification
system. This fragment was cloned into the Sal I digested, calf
intestinal phosphatase-treated (Gibco BRL) Baculovirus
expression vector Fast Bac HTb (Gibco BRL) The phosphatase
was used to remove 3' and 5' phosphates on the vector to lower
background by markedly reducing vector self ligation. Ligation
was done with T4 ligase (Promega) according to standard
protocol. Then 1-1 Ong of ligated DNA's were used to transform
DH 5 alpha E-coli subcloning cells (Gibco BRL) and recombinant
clones screened by mini-prep isolation and digestion with Sal I
restriction enzyme to determine the presence of p70 insert DNA.
Conformational restriction analysis was done with several other
restriction enzymes to determine proper oreintation. The T229A
and WT deletion 104 clones were doned in a similiar manner. The
initial DNA was subcloned from previously constructed pXST
clones ( Sugitama et al ., 1996). The T229A mutant was
generated by PCR specific primers that incorporates Sph I and Nsi
I sites at the T to A transition and uses the wild type p70 clone
as a template. This cDNA clone was digested with Sal I and Bal I
to isolate by agarose gel the T229A mutant fragment and
subcloned into the HTb baculovirus vector using a Nhe I linker.
The vector was treated with Sal I and Xba I enzymes, Xba I being
compatible with Nhe I. Ligaton at the Sal land Nhe MXba Isite
creates the completed clone. The c-terminal deletion 104 clone
30


was made by direct subcloning from pXST vector to the fastbac
HTb vector by the Sal land Xba I sites. Multi enzyme digestion
was done on these clones as well. Once positive clones were
identified, large scale maxi-prep (Gibco-BRL) was done to obtain
large quantities of DNA for further experimentation.
Transformation and Transposition of BAC p70 Clones into BAC1Q
Cells for Integration of p70 S6 Kinase into the Baculovirus
Genome
The three p70 clones were then transposed (site-specific)
into the baculovirus genome via the transformation of DH10 Bac
competent bmon14272 ( Gibco) E. coli. cells. These cells contain
low copy number mini-F replicon, a kanamycin resistance marker,
and a lacZ alpha coding region. In the N-terminus of lacZ protein
is an attachment site for the bacterial transposon Tn7 which does
not disrupt the reading frame of lacZ. The native Bacmid
propogates in the E. coli. Bac 10 cells and complements the lacZ
deletion on the chromosome to form colonies that are blue. The
recombinant p70 bacmids contain the P70 DNA in the mini-Tn7
site which when transposed into the viral genome disrupts the
expression of Lac Z alpha peptide making white colonies. For each
transposition, 1ng of DNA was used in 5|il of water and added to
100ml of Bac10 cells on ice for 30 minutes. The mixture was
incubated in a 42 C water bath for 45 seconds to heat shock and
31


then back in ice for two minutes. Then 900 jj,I of S.O.C. medium
was added and the mixture was incubated for 4 hours at 37 C.
Serial dilutions were plated on Luria agar plates containing 50
mg/ml kanamycin, 7 mg/ml gentamycin, 10mg/ml tetracycline,
100mg/ml Blue-gal and 40mg/ml IPTG. Plates were incubated for
at least 24 hours at 37 C to observe colonies. Positive white
colonies were restreaked and a well isolated colony was used for
liquid culture for the isolation of high molecular weight
recombinant viral DNA.
Isolation and Analysis of Recombinant p70 S6 Kinase Bacmid DNA
for Transfection and Viral Production
Isolated colonies were sterily transfered to 2ml of LB
medium containing 50mg/ml kanamycin, 7mg/ml gentamycin, and
10 mg/ml tetracycline and grown at 37 C for 24 hours shaking
at 300 rpm. Cells were centrifuged at 14000 x g for 1 minute and
resuspended in 0.3 ml of solution 1 [15mM Tris-HCI (ph8.0), 10mM
EDTA, 100mg/ml RNase A] and then 0.3 ml of solution 2 (0.2N
NaOH, 1% SDS). 0.3 ml of 3M potassium acetate was added and
samples spun at 1400x g for 10 minutes. DNA in supernatant was
precipitated in 0.8 mis of isopropanol for 10 minutes and
centrifuged again for 15 minutes, washed with 70% ethanol, and
the DNA pellet air dried for 10 minutes. The DNA was resupended
in 40|il of TE for further analysis. Five ml of Bacmid DNA was
32


analyzed on a 0.5% agarose gel containing 0.5 mg/ml of ethidium
bromide in TAE buffer. Samples were run at 23 volts for 12-15
hours to identify the 135 kb viral genome. Samples positive for
Bacmid DNA were used for transfection.
T ransfections
Transfections were carried out by seeding 9 x 10^ cells in a
6-well plate containing SF-900 SFM media with
penicillin/streptomycin at 0.5X final concentration for 2 hours to
attach to the well. Then 5p,l of bacmid in 100ml SF-900II SFM per
transfectant was mixed with 100 p,l of SF-900II SFM containing 6
ml of CellFectin reagent (Gibco) for 45 minutes at room
temperature. The lipid/DNA complexes (diluted tolml) was
overlayed onto the insect cells in the 6-well plates seeded
earlier, each containing 0.8 mis of SF 900II SFM. The complexes
were allowed to enter the insect cells by incubating for 5 hours
at 30 C. Then 2 ml of SF900II SFM was added to each plate and
incubated for 48 to 72 hours. Protein expression and activity was
analyzed after 48 hours and virus particles were harvested after
72 hours. Viral titers were determined by viral plaque assay.
Briefly, 6 x 105 insect cells were seeded into 6-well plates for
2 hours to make a unifrom monolayer. An 8 log serial dilution
was made of the virus harvested from the 72 hour transfections
33


and 1ml was layered onto the monolayer on insect cells for 1hour.
A 4% agarose solution was set at 70 C and 1.3 X SF900II insect
medium was placed at 40 C for making the SF-900 plaquing
overlay solution. Mixed 30ml of 1.3X insect media with 10 ml of
the agarose solution and placed at 40 C. Sequentially, removed
viral supernatants form cells and replaced with 2 ml of diluted
agrarose/media solution. Agarose was allowed to harden for 20
minutes and the plates were incubated at 30 C for 4-14 days.
Viral plaques were visible after 5-7 days and at the proper
dilution, 3-20 plaques were counted after no changes in plaque
number was observed. Typically, viral titers were 1-2 X 10
pfu/ml. Viral titers were amplified by sequential infections to
achieve 1 X 10 pfu/ml.
Infections
Viral infections were done with 0.5-1.0 ml of transfection
supernatants in 30 mis of insect cells (2 X 10 cells/flask) for
48-72 hours. Cells were harvested and lysed in Ripa buffer or
p70 buffer and the 6 HIS tagged p70 was precipitated by Talon
metal affinity resin (Clontech Laboratories, San Diego,CA).
Typically, 20 pi of Talon resin per 500ml of cel lysate roated at
4C for 1 hour and washed 3 times with lysis buffer. Samples
were then boiled for 5 minutes in 2X gel loading dye and analyzed
34


on 7.5% SDS-PAGE. Alternatively, the Talon resin/p70S6K protein
complex was used in an invitro kinase assay to determine activity
S6K
of p70 on S6 peptide. The complex was washed in kinase
buffer(10mM Hepes buffer, 1mM MgCI2, BSA, IP-20 1mM
DTT). To each sample was added 1X kinase buffer containing
100|iM nonradioactive ATP, 100 jim S6 peptide, and 200 p.Ci/ml
y32P ATP. Samples were incubated for 15 minutes at 30 C and
the reaction stopped with 20 |il of 0.5M EDTA. Then 25 |il of the
reaction mixture was loaded in duplicate on p81 phosphocellulose
paper and washed 5 times with 1/200 dilution of phosphoric acid.
The papers were dryed under IR light and counted in a liquid
scintillation counter.
CDK1/Cvclin B Phosphorylation of Recombinant p70 S6 Kinase
Site specific phosphorylation of p70S6K by CDK1/cyclin B
complex was investigated by invitro kinase assay, and
phosphorylated proteins subjected to tyrptic peptide mapping.
The recombinant p70 proteins made by baculovirus expression
system were employed as whole protein substrates to show
differential phosphorylation by CDK1/cyclin B as compared to
CDK2/cyclin E. Initially, 1 X 103 insect cells were infected with
35


either wild type, T229A mutant, or wild type deletion mutant for
S6K
48 hours, harvested, lysed in p70 lysis buffer, and the
supernatants transferred to a new tube containing 0.1 ml of Talon
purification resin. Samples were rotated at 4C for 1-2 hours.
Protein complexes were centrifuged at 14,000 X g for 1 minute
and washed with 1 ml of lysis buffer 3 times. Then the complexes
were resuspended in 50 |il of CDK kinase buffer containing 0.1 mM
ATP, 200 pCi of y^P ATP, and 1 jllI of CDK1/cyclin B complex
(MBP) or 1 pi of CDK2/cyclin E. The reaction proceeded for 30
minutes and was stopped by the addition of 20 pi of 0.5 M EDTA.
Then 50-60ml of loading dye was added and the samples were
boiled for 5 minutes. After brief centrifugation, the samples
were loaded onto a 10% SDS-PAGE and electrophoresed. The gel
was coomasie stained, dried, and autoradiographed.
Phosphooeptide mapping
Protein Precipitation and Digestion. Proteins were cut out
of SDS-PAGE gels and suspended in 50 mM ammonium bicarbonate
containing 1% SDS and b-mercaptoethanol and homogenized with
microcentrifuge dounce (Kontes). Samples were boiled for 5
minutes and rotated at room temperature overnight. Proteins
eluted out of the gel were precipitated with 10 mg of RNAse
carrier protein and 95% trichloroacetic acid for 1-2 hours.
36


Proteins were pelleted by centrifugation at 1400 X g for 10
minutes, supernant drained and pellet washed with 0.4 ml of
equal volume ethanol and ether and then dried for 5-10 minutes in
a speedvac. Samples were treated with 20 |il of formic acid/
H2O2 (10:1) for 1-2 hours at 4 C and then 0.4 ml of water was
added and the samples dried in a speedvac. Proteins were
subjected to Chymotrypsin and Trypsin digestion at 1 mg/ml in
NH4HCO3 at 37 C for 24-48 hours each. Following digestion,
0.4ml of water was added and samples were heated at 65 C for
10 minutes, centrifuged for 10 minutes, supernant transfered to a
new tube and dried completely. Samples were washed again with
0.4 ml of water two more times and dried. Samples were
subjected to Thin layer electrophoresis and Thin layer
chromatography.
Electrophoresis/Chromatoaraphv. Phospho-peptide samples
were dissolved in 10ptl of electrophoresis buffer ( 2.25% Formic
acid, 7.75% acetic acid) and sepaprated by charge in one
dimension using the Hunter thin layer peptide mapping system
(HTLE-7000, CBS Scientific). For each sample, 5-10 jxl was
loaded onto TLC-cellulose plates (Merk5716 celluose, 20X20 cm,
EM Science) using a blow dryer to concentrate the sample. The
plate was pre-wetted with electrophoresis buffer using a soaked
37


Whatman paper. The samples were electrophoresed for 30
minutes at 1200 volts and plates dried for 1 hour before thin-
layer chromatography. The plates were placed in a TLC chamber,
in an ascending position, containing chromatography buffer (65%
isobutyric acid, 5% pyridine, 3% acetic acid, and 2% butanol) and
allowed to run for 16-20 hours or until solvent front reached the
top of the plate. Plates were then dried for 3 hours and
autoradiographed.
Synthetic p70 S6 Kinase Ser 411 Phospho-Peptide Map. The
p70S6K peptide EPKIRSPRRFIG (Residues 406-417 of human p70S6K),
including the serine 411 phosphorylation site, was synthesized to
confirm migration of the Ser411 peptide during mapping. The
peptide was labelled invitro by CDK1/cyclin B kinase reaction
(incorporation of y32p ATP) and the resulting phosphopeptide
purified by C-18 Sep-Pak cartridge ( Millipore, Milford, MA),
which removes unincorporated y32pATP. Briefly, reaction mixture
was loaded onto a prewashed column and then washed extensively
with PBS until no 32p activity could be detected in the wash. The
phosphopeptide was eluted out with 70% acetonitrile in a total
volume of 0.1ml and dried in a speedvac. Then 40 pi of trypsin
was added and digestion proceded for 24 hours. The sample was
washed 3 times in water, drying in a speedvac each time. After
38


final wash, the peptide was resuspended in 0.1 ml of TLE buffer
and subjected to the TLE/TLC procedure.
CDK/Cvclin In Vitro Kinase Assay
Specific activities of cdks/cyclins complexes were
determined by (32p) incorporation into pH1 or p70CT peptide in
the immune complex. Substrate peptides used here were; pH1,
AVAAKKSPKKAKKPA (Residues 139-153 of trout Histone H1), pRB,
RPPTLSPIPHIPR, and p70 CT, EPKIRSPRRFIG (Residues 406-417 of
S6K
human p70 ). The peptides were prepared by automated solid-
phase peptide synthesis using a type 9050 automated synthesizer
from MilliGen/Biosearch, and the purity and concentration of the
peptides were confirmed by HPLC. A purified Xenopus
CDK1/cyclin B or a CDK2/cyclin E complex was prepared as
described previously. Cells (5 x 10) were washed with PBS and
lysed at 4 C in 1ml of lysis buffer (50mM Tris-HCL, pH7.4, 1mM
EDTA, 25mM NaCI, 40mg/ml PMSF and 0.1% np40).
The extract (500p.l ) was incubated for 1 hour at 4 C with either
anti-CDK1, CDK2, cyclin A, or cyclin B antibody. The immune
complex was absorbed to Protein G- coupled sepharose beads for
1 hour. Alternatively, the extract was incubated for 1 hour at 4
C with p13 sue agarose beads. The beads were washed three
times with lysis buffer, and once with kinase buffer (50mM Tris-
39


Hcl, pH7.4, 10mM MgC^, 1mM dithiothreitol). Following the final
wash, the immune complexes were suspended in 50 p,l of the
kinase buffer containing 100mM unlabeled ATP, 200 mCi/ml
|-y32p] atP, ancj -)q4 y peptide. The reaction was allowed to
proceed for 15 minutes at 30C and terminated by the addition of
1Opl of 500 mM EDTA. Following a brief centrifugation, the
supernatant (20 (il) was applied to P-81 phosphocellulose paper
(Whatman 3698-025) and radioactivity was determined using a
liquid scintillation counter (Bio-Safe II, Research Products Inc.,
Mount Prospect, Illinois). In some experiments, a purified
Xenopus CDK1/cyclin B complex or a purified Xenopus CDK2/cyclin
E complex was used instead of immune complexes. Kinetic
constants were determined using 10' to 10" M peptide.
T cell Time Course
T Cell CT-p70 S6 Kinase In Vivo Phosphorylation. T Cells were
stimulated with PDB / lonomycin and harvested at 3,24,39,and 48
hours post-activation with and without rapamycin. Each time
point was labelled in phosphate free dMEM containing 1mCi of ^P
inorganic phosphorous (NEN-Dupont, 1mCi/ml, Nex-053, Boston,
MA) for 3 hours prior to harvest. Cell pellets were lysed in Ripa
buffer for 15 minutes. p70 S6K was immunoprecipitated by anti-
CT p70 antibody for 24 hours and proteins analyzed by 7.5%
40


sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-
PAGE), Coomassie stained, and autoradiographed.
Activity Time Course of p70 S6 Kinase and CDK1. T cells
harvested at the same time points were lysed in p70S6K lysis
buffer( 10mM KP04, 1mM EDTA, 5mM EGTA, 10mM MgCI2, 50mM b
Glycerol phosphate, 1mM Na3V04, 2mM dTT, 40mg/ml PMSF, 0.1%
NP-40) or CDK1 lysis buffer ( 50mM Tris pH7.4, 5mM EDTA,
250mM NaCI, 0.1% NP-40). CDK1 and p70S6K was
immunoprecipitated for 1 hour by anti-CDK1 antibody and anti-CT
p70S6K antibody respectively,and complexed to protein G sepharose
S6K
beads for 1 hour by rotating at 4C. For p70 activity, the beads
were then washed twice in lysis buffer and once in kinase buffer
(20mM MgCI2, 40mM Tris pH7.5, 2mg/ml IP-20, 0.2mg/ml BSA,
0.8mM dTT). Then 50(i.l of kinase buffer was added containing
0.1 mM ATP, 10mM S6 peptide, and 10jnCi (10|iCi/ml) of y^2P ATP
(NEN-Dupant, 30Ci/mMol) and reaction set at 30 C for 15
minutes. After brief spin, 26 \i\ of the reaction supernant was
loaded in duplicate on p81 phospho-cellulose paper disks
(Whatman 3698-025), washed in 0.5% phosphoric acid five times,
and dried under infrared lamp for 15 minutes. The disks were
counted in a liquid scintillation counter. For CDK1 activity, the
beads were washed twice in lysis buffer and once in CDK1 kinase
buffer( 100 mM Tris pH7.4, 20mM MgCI2, 2mM dTT). Kinase
41


reaction procedure was completed similiarly using H1 peptide as
a substrate.
Cell Cycle FACS Analysis. T cells collected at the same time
points were analyzed to determine the percentage of cells in each
phase of the cell cycle. Cells were harvested, washed in PBS,
resuspended in 0.2ml of saline, and transfered to a 12X75mm tube
containing 3ml of 70% ethanol. Samples were stored at -20C
overnight, centrifuged and washed with PBS. The fixed cells were
incubated with 0.5ml of 0.25mg/ml ribonuclease (Sigma) in PBS
at 37 C for 10 minutes. The cell suspension was mixed with 0.5
ml of propidium iodide (Calbiochem, La Jolla, CA) solution
(50mg/ml in PBS), and after 60 minutes analyzed by a flow
cytometry (EPICS Profile, Coulter) collecting red fluorescence
(>600 nm) with 488 nm excitation.
Increase of CT-p70 S6 Kinase Phosphorylation bv Nocordizole
Treatment In Vivo
Cos cells at 30% confluency were set up in dMEM media and
transfected with 10 mg of pCMV5-WTp70 plasmid using 30 ml
of cell fectin reagent (Gibco-BRL) for each plate. Cells were
incubated for 40 hours and then pretreated with either
Nocordizole (Sigma) at 500 mm, Aphidicolin (Sigma) at 1 mm, or
Rapamycin at 1 mm for 5 hours. Then cells were incubated for 3
42


hours in the presence of 32P orthophosphate using dMEM without
Na phosphate supplemented with 10% dialzed FCS. Cells were
harvested and lysed in p70 lysis buffer for 15 minutes,
centrifuged and the supernatant transfered to a new tube. Anti-
CT-p70S6K antibody (0.5 mg per IP) was added to the lysates and
incubated on ice for 30 minutes followed by 30 minute incubation
with 30 pil sepharose beads. Immune complexes were washed 3
times with p70 lysis buffer and 40% of the beads were boiled in
loading dye and run on a 7.5% SDS-PAGE, the gel fixed and dried
and autoradiographed.
43


CHAPTER 3
Results
The studies presented here investigate the possible
connection between the cell cycle regulatory protein CDK1 and
the major physiological kinase responsible for phosphorylating
the S6 protein, referred to as p70 S6 Kinase. Located in the
carboxy-terminal region of the kinase are four serine/threonine
phosphorylation sites, each of which are followed by a proline
residue. This represents a commonly used substrate recognition
motif for cyclin-dependent and mitogen-activated protein
kinases. An earlier study showed that CDK1 could indeed
phosphorylate C-terminal p70S6K in vitro. To further analyze this
phenomenon, it was determined if other cyclin-dependent kinases
could also use p70 S6 kinase as a substrate or was this specific
to CDK1 alone. Among the CDKs identified to date, CDK2 is the
closest kinase to CDK1 by homology and substrate specificity.
Therefore I initially compared CDK2 and CDK1 in the ability to
S6K
phosphorylate recombinant p70 I developed a system to study
S6 K
specific phosphorylation events on the carboxy-terminus of p70
44


to truly analyze the activity of cell cycle regulatory proteins on
this enzyme.
Baculovirus Expression System Provides High Quantities of
Functionally Active Wild-Tvpe or Inactive Mutant p70 S6 Kinase
for Phosphorylation Analysis
In order to study phosphorylations at specific sites on
S6K
p70 , there was a need for large quantities of native proteins
that could be readily phosphorylated by various enzymes and
digested with proteinases for tyrptic phosphopeptide analysis.
The Baculovirus expression system provided an excellent way to
accomplish these things because of the availability of highly
efficient vectors for production of high quantities of
heterologous proteins. The recombinant proteins contained a
region upstream of the coding region consisting of six histidines
which allowed for easy purification by conjugation to metal
resin agarose beads. Phosphorylation of p70S6K was carried out
directly on the resin beads. Two mutants were constructed which
enabled analysis of the two sets of phosphorylations sites on p70
S6K
. T229 A is a mutation in group 1 and a carboxy-terminal
deletion mutation was also used to verify the four CT-
45


phosphorylation sites of group 2. Figure 5 shows the isolation
and characterization of the positive clones using Baculovirus
expression vector Fastbac Htb and the restriction analysis
confirming positive oreintation and correct reading frame of the
transposing DNA (Figure 5a). Following transposition, Bacmid
DNA was isolated which contained the cloned p70 of interest
(Figure 5b). The subsequent expression levels of the three
variants of p70 were all very high (Figure 5c) and migrated at the
correct sizes. These proteins were analyzed for kinase activity
and only the wild type had good activity, with the deletion mutant
having partial activity and the T229A mutant nearly inactive
(Figure 5d). As will be seen later, these recombinant proteins
provided excellent substrates to determine differences in CDK1
and CDK2 phosphorylation of CT-p70S6K.
46


A)
B)
g rec T229A-p70 rec ACT-p70
kb ^ 1 2 3 4 5 6 1 2 3 4,5 6
D)
soooo-i
g 4SOCO -
40000 -
>- 3S000-
> 30000-
^ 29000
o 1SDC0-
10000-
sxo- HH
0Lmmm________INK________HH________HBBL
BKG WT T229A WTDEL
S6K
Figure 5. Characterization of the recombinant p70 clones generated by the Bac to Bac
Baculovirus expression system. (A) Restriction analysis of the cDNA clones after fast Bac
HTB vector ligation with p70S6K cDNA showing positive orientation. (B) Isolation of
recombinant Bacmid DNA from T229A and D WT clones after transposition event with
S6K
vector constructs. (C) Western Blot showing expression of P70 proteins generated by
Baculovirus system. (D) The kinase activities of the three enzymes, as determined by
phosphorylation of the S6 peptide. See text for details.
47


CDK1/Cvclin B and Not CDK2/Cyclin E Phosphorylates
Recombinant CT-p70 S6 Kinase on Serine 411
Phosphopeptide mapping analysis and baculovirus generated
p70 recombinant proteins provided the tools necessary to
demonstrate the specificity of CDK1/cyclin B phosphorylation on
serine 411 of CT-p70S6K. The rec WT-p70, rec T229A-p70, and
the rec DCTp70 proteins generated by viral infection of insect
cells, were compiexed with Talon metal resin beads and then used
as substrates for either CDK1/cyclin B (Xenopus maturation
promoting factor) or CDK2/cyclin E in vitro kinase reactions.
Histone H1 was also a substrate as a control of phosphorylation.
The recWT-p70 protein was phosphorylated at the same intensity
with and without CDK1/cyclin B (MPF) due to a high level of
S6K
autophosphorylation (Figure 6a). Therefore the recT229A-p70
was used as a substrate for all further phosphorylations because
the group 1 T229A mutation significantly reduces the
autophosphorylation of the enzyme.
48


A
B
co
o
*
D
O
recWT-p70
Histone H1 recT229A-p70 recACT-p70
LLJ CO LU CO LU CO
C c C c c c
75 75 73 p "o >s £
o O o o o o
e c o Q 52 Q 2 E o cvr Q T~ Q o u c o * o T a
o O O O O O o o o
recT229A-p70
recACT-p7G
Histone H1
Figure 6. One-dimensional SDS-PAGE showing in vitro phosphorylation of recombinant
p70S6K proteins by CDK 1/cyclin B and CDK2/cyclin E. (A) The phosphorylation of wild-type
p70S6K with and without CDK1/cyclin B. (B) The phosphorylation of p70S6K mutants with
CDK1/cyclin B or CDK2/cyclin E. Histone H1 as a control of phosphorylation ability. See
text for details.
49


In one dimensional 7.5% SDS-PAGE, CDK1/cyclin B
phosphorylates rec-T229A p70S6K more intensively than
CDK2/cyclin E which shows only background autophosphorylation
(Figure 6b). Both CDK complexes exhibit good kinase activity as
seen by the intense phosphorylation of Histone H1. As a
comparison, rec DCTp70 showed only very slight
phosphorylation without the carboxy-terminal group 2
phosphorylation sites. To further examine which serine or
threonine sites on p70S6K are phosphorylated, the three bands from
the recT229A 7.5% gel were cut out and subjected to trypsin and
chymotrypsin digestion and the resulting phosphopeptides
analyzed by the two-dimensional mapping procedure. Figure 7
shows the phosphopeptide mapping results of these three labeled
S6K
bands. Neither p70 itself nor CDK2/cyclin E phosphorylated
p70 show any specific phosphorylations with the apparent
absence of carboxy-terminal phosphopeptides (Figure 7, panel A
and B). CDK1/cyclin B however, exhibits good phosphorylation of
the carboxy-terminal sites with one spot showing the very
intense labeling (Figure 7, panel C). This spot was assumed to be
the Ser 411 phosphorylation site based on the presence of the
CDK1 consensus motif at that position. The serine 411
phosphorylation site (EPKIRSPRRFIG) was confirmed by the
migrational analysis of synthetically made peptide representing
this serine 411 peptide sequence. This phosphopeptide migrated
50


to the identical position on the cellulose plate (Figure7, panel D).
Taken together, these data demonstrate that among the C-
terminal phosphorylation sites, the serine 411 on p70S6 kinase
is most highly phosphorylated by CDK1/cyclin B in vitro.
51


>rec7229A-p70
Control


rec7229A-p70
CDK2/Cyc!ir, E
! B
rec 1229A-p70
CDKVCyclin B .

-/ .- .:.. .::-j -,'v->-;'.
mKMKmKKMKHKHi
', ' ., s
:#
WtmyM&MWim,
as
.,* v *
[C
EPKIRSPRRHG
CDK1/Cydin B
D
*


Figure 7. Two-dimensional thin-layer electrophoresis of the in vitro
phosphorylated bands of T229A-p70S6K cut out from the one-dimensional gel seen in
figure 6. Phosphorylated T229A-p70S6K proteins were subjected to chymotrypsin
and trypsin digestion and resulting phospho-peptides separated by charge during
thin-layer electrophoresis and then by size during thin-layer chromatography. (A)
The T229A-p70S6K protein without CDK1/cyclin B added. (B) The phospho-
peptides from CDK 2/cyclin E phosphorylated T229A-p70S6K. (C) Separation of
phospho-peptides from CDK1/cyclin B phosphorylated T229A-p70S6K. (D) The
confirmation of the Ser 411 carboxy-terminal phosphorylation site by the similar
migration of synthetically prepared Ser 411 peptide after in vitro phosphorylation
by CDK1/cyclin B. See text.
53


To confirm that these phosphopeptides really represent the
C-terminal phosphorylation sites, cos cells were transfected
S6K S6K
with WT-p70 or DCT-p70 (cloned into mammalian
transfection vector pCMV5), labelled with 32P inorganic
phosphorous and subjected to the same tryptic peptide mapping
methodology. Figure 8 shows the phosphopeptide maps of p70 S6K
and the individual phosphorylation sites. Panel A of figure 8
shows the migration of each phosphopeptide and the directions of
TLE and TLC. The spots designated as C correspond to the
carboxy-terminal sites (Ser411, Ser418, Thr421, Ser424) which
were completely absent in the carboxy-terminal deletion mutant
(Figure 8, panel B). Thus this methodology provides a suitable
S6 K
way to distinguish specific phosphorylations on the p70
molecule.
54


TLC
A= T229
rec Ap70S6K


Figure 8. Characterization of the chymotrypsin and trypsin generated phospho-
peptides of p70S6K. The sites marked A and B were determined to be T229 and T389
of set 1 phosphorylation sites. The C-terminal phosphorylation sites of set two are
indicated by the drawn circle in the upper right quadrant of the map. The
confirmation of the C-terminal phosphorylation sites is shown by the use of the C-
terminal deletion mutant of WT-p70S6K, which shows the complete absence of the
three carboxy-terminal phosphorylation spots
56


The Rate of Peptide Phosphorylation by CDK1/Cvclin B Complex is
Higher Than CDK2/Cvclin E Complex
In agreement with my results, the region surrounding ser
411 has the best match among the C-terminal phosphorylation
sites in p70S6K with the consensus substrate sequence of CDK1
(K/R-S-P-R/P-R/K/H) determined by a peptide selection
approach.
To further analyze the differential phosphorylation activity of
CDK1 and CDK2, CTp70S6K peptide(EPKIRSPRRFIG) or histone H1
peptide(AVAAKKSPKKAKKPA) was made synthetically by amino
acid synthesizer and assayed for relative rate of phosphorylation
by either purified Xenopus CDK1/cyclin B complex or CDK2/cyclin
E complex. Kinase reactions containing CDK complexes, y 32P
ATP, and 10"3 M to 10'3 M concentration of peptides were
incubated for 15 minutes at 30 C. Figure 9 shows the
differential enzyme kinetics of the two cyclin-dependent kinase
complexes. CDK1/cyclin B complex has over ten fold higher rate
of phosphorylation on CT-p70S6K peptide, with an apparent Km of
50|iM, than does CDK2/cyclin E complex (Km 500|iM). This was in
contrast to the rates of phosphorylation on pH1 (Km: for CDK1 ~10
mm, for CDK2 ~20mm). The specific activities were divided by
57


the activity to phosphorylate 10"^ M of pH1 peptide to normalize.
To further demonstrate the specificity of this CDK1
phosphorylation event, I also analyzed the early cell cycle
proteins CDK4 and CDK6 for their potential ability to
S6K
phosphorylate the C-terminus of p70 . Figure 10a shows the
rate of phosphorylation of CDK 4 and CDK 6 on p70S6K as compared
to CDK1. CDK 4 and CDK 6 both show very little phosphorylation
of p70 however, figure 10b shows that the Immunoprecipated
CDK 4 and CDK 6 proteins are active as seen by their ability to
\
phosphorylate their known substrate, the whole retinoblastoma
protein (Rb).
58


concentrations of p70CT (M)
Fig. 9. Relative phosphorylation rate of p70S6K by purified CDK1/cyclin B
or CDK2/cyclin E. A 10"8 to 10'3 molar range of C-terminal p70S6K peptide
was used to assay activity of purified Xenopus CDK1/cyclin B (apparent Km of
50 pm) or CDK2/cyclin E (apparent Km of 500 pm).
59


Phosphorylation rate of p70CT
Phosphorylation of pRb
B
-3
8000
E
Q.
O
Figure 10. Relative rate of phosphorylation on p70S6K C-terminal peptide by
immunoprecipitated CDK4 and CDK6 from cycling Jurkat cells (A). A 10-5 to 10-3
molar range of substate was used, and CDK1 was used as a control. The
phosphorylation activity on known substate pRB is also shown to confirm that CDK1,
CDK4 and CDK6 are active complexes.
60


CDK1/Cvclin B Preferentially Phosphorvlates CT-p7Q S6 Kinase
Whereas Other Members of the Cvclin-Dependent Kinase Family Do
Not
Since there was a significant difference in rate of
phosphorylation using purified CDK complexes, the next question
was to ascertain phosphorylation rates of CT-p70 by other
members of the cyclin-dependent kinase family directly
immunoprecipitated from cycling T lymphocytes. Lysates were
made from T cells stimulated with PDB/ionomycin for 48 hours
and treated with antibodies against CDK1, CDK2, CDK4, CDK6,
cyclin A, and cyclin B. Lysates were also treated with p13suc-
agarose beads which immunoprecipitate both CDK1 and CDK2 and
their associated cyclins. These immune complexes were
conjugated to sepharose beads and assayed for phosphorylation
S6K
activity on CT-p70 peptide EPKIRSPRRFIG by in vitro kinase
assay. Phosphorylation activity of pH1 peptide and pRb peptide by
these immune complexes served as a control. Table 1 shows the
relative phosphorylation rates of the synthetic peptides using the
various antibodies. CDK1 and cyclin B more effectively
phosphorylated CTp70 peptide than the other members of the
CDK family. p13suc beads also demonstrated significant activity
61


due to the presence of CDK1 in the complex. The data is presented
as a percent activity to the specific activity to phosphorylate
pH1 for CDK1 and CDK2; whereas the percent activity to
phosphorylate pRb was used to normalize for CDK4 and CDK6. It
should be remembered that CDK1 can associate with either cyclin
A or cyclin B and that CDK2 can associate with either cyclin A or
cyclin E, but only CDK1/cyclin B had significant activity in theses
lysates. It is unlikely that CDK1/cyclin A complex
S6K
phosphorylates CT-p70 but it is not clear because the ratio of
CDK1 and CDK 2 in cyclin A complexes was not determined here.
Both CDK4 and CDK6 associate with the D type cyclins. Thus,
these data add supporting evidence that there is specific
differential phosphorylation activity on the carboxy-terminus of
S6K
p70 by cell cycle regulatory proteins in activated T
lymphocytes.
Table 1. Relative phosphorylation rate of synthetic peptides by CDK1-,
CDK2-, cyclin A-, cyclin B-, or pl3suc-immune complexes.
antibodies relative phosphorylation (%) rate
pHl p70CT
anti-CDKl 100 56.3 14.2
anti-CDK2 100 5.4 4.0
anti-cyclin A 100 4.4 0.3
anti-cyclin B 100 42.5 12.0
p!3suc 100 45.9 10.9
62


Differential CDK1 Activity Phosphorvlates p70 S6 Kinase C-
Terminal Peptide at a Different Cell Cycle Phase Than Histone H1
Peptide
Next we investigated where in the cell cycle the carboxy-
terminus of p70S6K is phosphorylated by CDK1/cyclin B. Based on
the results of Table 1, p13suc beads had significant kinase
activity to phosphorylate CT-p70 peptide due to sufficient
CDK1/cyclin B activity. Therefore p13suc beads were employed
to explore the phosphorylational status of CDK1 on the carboxy-
terminus of p70S6K peptide as compared to Histone H1 peptide
during cell cycle progression. The p13suc is a CDK-associated
protein which binds to CDK1 and CDK2 and possibly CDK3 and
precipitate them in an active state. Lysates from human primary
T cells stimulated with PDB/ionomycin were collected at
selected time points of the cell division cycle and incubated with
p13suc agarose beads for 1 hour and subsequently assayed for
kinase activity on either histone H1 peptide or the CT-p70S6K
peptide. As seen in figure 11, the activity to phosphorylate the
KKSPKK peptide derived from Histone H1 was increased at about
the time when CDK2 was activated in the cell cycle corresponding
to the G1/S phase. In contrast, the activity to phosphorylate the
KIRSPRR peptide derived from CT-p70S6 kinase was increased
later in the cell cycle, rather in parallel to CDK1 activation at the
G2/M boundary. These results suggested that the cyclin-
63


S6K
dependent kinase activity on CT-p70 was distinct from the
histone H1 phosphorylation earlier in the cell cycle.
T cells stimulated with
PDBu/lonomycin
Fig. 11 pl3suc-associated kinase activities to phosphorylate pHl o r
C-terminal P70S6K peptide. Stimulated T lymphocytes were harvested
at sequential times throughout the cell cycle. CDKl/cyclin B was
immunoprecipitated using pl3suc beads and the kinase activities of
each time point assayed with either pHl or p70CT as a substrate. FACS
analysis was done to confirm cell cycle timing.
64


In summary, the in vitro studies presented evidence of
differential cyclin-dependent kinase activity on a specific
peptide sequence located in the carboxy-terminus of p70S6K, that
demonstrated over 10 fold higher affinity for CDK1/cyclin B than
CDK2/cyclin E. This was shown to be a phenomenon more
exclusively associated with CDK1 than other members of the
cyclin-dependent kinase family such as CDK2, 3, 4, and 6; and this
phosphorylation activity correlated with the G2\M phase of the
cell cycle, distinct from the more common substrate histone H1
which becomes phosphorylated around the G1/S transition. These
in vitro findings established the rationale to search for evidence
that this was occurring in vivo and had physiological
significance. The focus shifted to exploring phosphorylation
events occurring in actively cycling cells in hopes of finding
further evidence linking components of the cell cycle regulatory
system to a key enzyme of an important cell signal transduction
pathway.
Late Cell Cycle Phosphorylation Versus Early Activity of p70 S6
Kinase
CDK1 is known to be highly abundant and active late at the
cell cycle in the G2 and M phases. In vitro evidence presented in
65


Figure 12 suggested CDK phosphorylation on CT-p70S6K peptide
was occurring late in the cell cycle at the G2/M boundary.
Therefore, I explored a potential link in the late cell-cycle
phosphorylation of the carboxy-terminal region of p70S6K and
components of the cell cycle machinery at the G2/M phase in
vivo. While investigating in vivo time course activity of p70 , it
became clear that their was a discrepancy in the early and rapid
activation of p70 in the group 1 phosphorylation sites and the
late cell cycle phosphorylation of carboxy-terminal group 2
phosphorylation sites. The activities of p70 and CDK1 as they
correlate to the cell cycle are shown in figure 12b and 12c. p70S6K
activity is most active very early after primary T cells become
activated by PDB/ionomycin and continually declines through
progression of the cell cycle. In contrast, CDK1 activity is
maximal very late in the cell division cycle corresponding to the
G2/M boundary (Figure 12c). The immunosuppresant drug
Rapamycin totally inhibits the p70S6K activity, however the late
activity of CDK1 is rather resistant to Rapamycin. FACS analysis
confirms cells were entering G2/ M phase only during the latter
parts of the time course (Figure 12d). In contrast to the early
S6K
activity of p70 , the enzyme became highly phosphorylated very
late in cell cycle progression in vivo. Figure 12a shows T cells
stimulated with PDB/ionomycin and labeled with inorganic ^2P
phosphorous for 3 hours prior to each harvest.
66


S6K
Immunoprecipitations with anti-CT p70 antibody show that p70
was intensively phosphorylated late in the cell cycle and that
this in vivo phosphorylation was largely resistant to Rapamycin
treatment indicating a possible connection to group 2
phosphorylation sites in the carboxy-terminus. At the protein
level, Figure 12a (second photo of Figure 12a) shows Western
blots indicating that p70 was indeed activated by PDB\I as seen
by the upper band at early time points, and this was inhibited by
the addition of Rapamycin. Total protein amount was similar for
each sample load.
67


PDBu/1
PDBu/l +RAP
(A)
24 39 48 3 24 39 48 hours
200
- P70S6K
p70S6K
- IgHC
(B)
(C)
(D)
E 50000
CL 40000
>. ** > 30000
s o (0 20000
it CD (0 10000
o r*. a 0
10000
E a o 8000
>. 6000
o 4000
<0 T- * 2000
a o 0
c 120 100
o +3 80
CB 3 60
a o a. 40
sP o'- 20
G0/G1
El S
G2/M


Fdure 12. The phosphorylation of p70S6K in cycling human T lymphocytes by 32P
o t lophosphate labeling in vivo. T cells were stimulated with PDB/lonomycin +/-
Rapamycin, radioactively labelled with 32P phosphorous and collected at selected time
nts throughout the cell cycle. Each time point was labelled for 3 hours. The p70S6K
pMteins were immunoprecipitated with anti-C-terminal p70S6K antibody, conjugated to
sepharose and separated on a 7.5% polyacrylamide gel. (A) The results showing 32P
lat elled p70S6K during the time course. Protein amount was nearly equal. (B) The
kijnase activity of the p70S6K present in each time point, assayed on the phosphorylation
o| S6 protein in vitro. (C) The kinase activity of CDK1 present in each time point,
assayed on the phosphorylation of Histone H1 protein in vitro. (D) FACS staining
showing the cell cycle timing of each sample, confirming that the late time point cells
were in the late phases of the cell cycle.
69


In a similiar experiment, T lymphocytes were treated with
Rapamycin only during 45-48 hours after stimulation with
PDB/lonomycin instead of stimulation from the beginning of the
time course. This much more specifically timed addition of
Rapamycin also showed no effect on the phosphorylation of p70S6K
as shown in figure 13. This reiterates the possibility that this
phosphorylation of p70S6K late in the cell cycle is principally due
to the phosphorylation occuring at the carboxy terminus. Protein
amounts for each lane were relatively constant.
70


Q.
<
Figure 13. The phosphorylation of p70S6K in vivo by 32P labeling in culture with and without
Rapamycin added specifically at 45-48 hours after PDB/lonomycin stimulation, as
compared to Rapamycin treatment at the beginning of the time course. Rapamycin also had
no effect when cells were treated at that time, suggesting the possibility that this
phosphorylation was the carboxy terminus of p70S6K. The lower photograph shows similar
protein amount for the PDB/I and PDB/I+RAP samples.
71


The Enhancement of CDK1 Protein Expression and Activity With
Nocordizole Treatment Also Markedly Increases d70 S6 Kinase
Phosphorylation In Vivo
The T cell time course experiments revealed a distinct late
cell cycle phosphorylation with potent CDK1 activity correlating
to the same phase of the cell cycle. But was CDK1 actually
phosphorylating CT-p70S6K in vivo? To provide some initial
supporting evidence I used ^2p labeled transfected cos cells to
study the phosphorylational status of p70 . The treatment of
these cells with known modulators of the cell cycle like
Nocodizole, Aphidicolin, and Rapamycin, made it possible to
inquire about specific p70 phosphorylations at selected times in
the cell cycle. Cos cells transfected with p70WT-CMV5 (a
mammalian transfection vector containing the sequence of wild-
type p70 cDNA) were pretreated 5 hours with the drugs prior to
harvest. Immune complexes constructed with CT-p70 antibody
S 6K
were made from these cells and assayed for p70
phosphorylational status. Nocodizole (an inhibitor of microtubule
assembly) stops the cell division cycle in the G2/M phase and
dramatically elevates the level of CDK1 in the cell (Figure 14).
Aphidicolin is an inhibitor of DNA polymerase alpha and delta, and
stops the cell cycle at the G1/S phase and only slightly altered
72


S6K
CDK1 activity. Rapamycin, a specific inhibitor of p70 had no
effect on CDK1 activity, however, it completely inhibited p70S6K
activity (Figure 14d). The activity of p70 was only moderately
affected by Nocodizole and Aphidicolin. With the dramatic
increase in CDK1 activity by Nocodizole treatment, the
phosphorylation of p70S6K was also increased while actual protein
amounts remain equal (Figure 14a and 14b). These results add
supporting evidence that CDK1 may, indeed, be involve in the
S6K
phosphorylation of p70 during the G2/M transition in vivo.
73


A
kD
B
C
con neo aph rap
P70S6K
IgHC
E
a
i*
>
*3
U
(0
JC
(O
0>
o
I*-
a
500000
400000
300000
200000
100000
0
z
a.
<
Q.
<
DC


Figure 14. Cos cells transfected with p70S6K WT and cultured in the presence of 32P
orthophosphate and either Nocordazole, Aphidicolin, or Rapamycin for 3 hours in
vivo. Cells were harvested and p70S6K immune complexes were analyzed by 7.5% gel
(A). Nocordazole increases p70S6K Rapamycin resistant phosphorylation. Western
blotting of the same gel to show equal protein amount (B). Each sample was analyzed
for P70s6K activity in vitro (C), and CDK1 activity in vitro(D). CDK1 activity is
very high in Nocordazole sample, whereas p70S6K activity does not correlate with
this in vivo phosphorylation.
75


CHAPTER 4
DISCUSSION
The investigations presented here demonstrate
phosphorylation activity on C-terminal p70S6K by CDK1 and not by
other members of the cell cycle regulatory machinery. This is
counterintuitive because p70S6K functions early in the cell cycle
during G1, and the CDK/cyclin pairs that should correlate with
p70 should be CDK4 and CDK6 complexes because they function
early in G1. This correlation should occur at the group 1
phosphorylaton sites. But substantial evidence is presented here
showing a specific phosphorylation on serine 411 in the group 2
phosphorylation site by CDK1/cyclin B in vitro. Indeed, in a recent
study of the substrate specificity of CDK1 and CDK2, KSPRK
sequence (closely matching Histone H1 peptide that I used) is
equally phosphorylated by CDK1/cyclin B, CDK1/cyclin A,
CDK2/cyclin A and CDK2/cyclin E (Holmes, J. K., and M. J. Solomon.
1996). In contrast however, when the last K amino acid is
substituted to R (making a peptide sequence very close to p70S6K
peptide I used), CDK2/cyclin A and CDK2/cyclin E do not
76


phosphorylate the peptide (<5%) whereas CDK1\ cyclin B
maintains about 70% of the activity to phosphorylate the original
peptide. This suggests that the R414 amino acid in the sequence
RSPRR of p70S6K is responsible for the preference of CDK1 as the
kinase to phosphorylate p70 . Given that CDK1 correlates with
the G2/M phase of the cell cycle, it is not surprising then that the
specific phosphorylation of CT-p70S6K by CDK1 occurs
predominantly late in the cell cycle. Preliminary evidence
presented here indicates that this phosphorylation event may be
occurring in vivo, and that this may be important for transition
into mitosis. The discrepancy in the early activity of p70 and
the late cell cycle phosphorylaton of the C-terminus indicates
two possible mechanisms regulating p70S6K; one Rapamycin-
sensitive and responsible for the enzyme's principal activity, and
one regulating another physiologically important interactions.
Implications for CDK1 Phosphorylating p70 S6 Kinase
Taken together theses findings present an interesting
question. What is CDK1/cyclin B doing functionally to the C-
terminus of p70 when it phosphorylates it and why is this event
necessary late in the cell cycle instead of early in G1? One
possibility could be that the kinase is rendered inactive by this
massive phosphorylation in the C-terminus. CDK1 is known to
77


phosphorylate and inactivate many proteins involved with
cytoskeletal reorganization during mitosis. However, we and
others have shown that when the C-terminus of p70S6K is mutated
or deleted, the kinase still exhibits activity due to the group 1
phosphorylation site in and around the catalytic site.
Alternatively, the kinase could be rendered constitiutively active
by altering the interaction between the pseudosubstrate in the
carboxy-terminus and the amino terminal acidic region. This
would alter the ability of the kinase to be activated correctly,
leading to an enzyme with the active site continually exposed,
and thereby leaving p70S6K in a hyperactive state until after
mitosis is complete. However in vivo data presented here
indicates p70S6K activity is seen only early after cell stimulation
with very low kinase activity at the time this C-terminal
phosphorylation occurs (see Figure 1b). Therefore, some other
function must be bestowed on the kinase to account for this
phosphorylation event.
Subcellular Localization
It can be speculated that there is a novel function for this
kinase late in the cell cycle elicited by the interaction of
S6K
specific binding proteins in and around the C-terminus of p70
Specific binding proteins altered by C-terminal phosphorylation
may in turn impart new physiological responses required for
78


division via differential subcellular localization of the kinase.
Preliminary immunofluorescent labeling studies indicate subtle
changes in subcellular localization of p70 in the G2/M phase,
with p70 co-localizing with actin filaments and some
differential staining showing a more perinuclear localization in
G2/M (unpublished observations). These binding proteins may
activate other biochemical pathways in the cytoplasm and
subsequently impart changes in the nucleus.
Nucleolar Organization
The nucleus in fact offers some other intriguing
speculation. Thomas et al. have demonstrated a distinct pool of
phosphorylated S6 in the nucleus. The absence of protein
synthesis in the nucleus suggests a unique role for this kinase
and this pool of nuclear phosphorylated S6. The nuclear isoform
of this kinase, p85 (containing the 23 aa nuclear localization
region in N-terminus), is expressed at a level 100 fold less than
p70 It could be speculated that the subcellular localization
changes induced by C-terminal phosphorylation of p70S6K may
affect or augment the phosphorylation of nuclear S6, which may
be involved in regulating ribosome biogenesis in the nucleolus.
During prophase of mitosis, the nucleolus first decreases in size
then dissociates. In metaphase, the nucleolus disappears
79


completely, and when ribosomal RNA synthesis restarts in
telophase, small nucleoli reappear at the chromosomal locations
of the ribosomal RNA genes (Miller, O.L. 1981; Anastassova-
Kristeva, M. 1977). It may be postulated that p70/85 plays an
important function when the nucleolus is disintegrated and
distributed over all metaphase chromosomes to be carried to each
of the new daughter cells and regenerated into new nucleoli.
Future Experiments
Several experiments will be conducted in our lab to address
some of these issues. Initially, I will further demonstrate the C-
terminal phosphorylation of p70S6K by CDK1\ cyclin B in vivo by
endoproteinase Glu-C (V8 protease) digestion of p70 from
actively cycling cells. The V8 protease isolates the specific C-
terminal region that was studied in vitro, so that this
phosphorylation event can be confirmed to be physiologically
important in the intact cell. Also I plan to construct a CDK1
dominant negative mutant and analyze the effects this may have
on the phosphorylation of C-terminal p70 in vivo to further
clarify the importance of this interaction. Additional subcellular
localization studies will be conducted with Nocodazole treated
cells to examine the effect of increasing the level of CDK1
expression on p70 function and its location in the cell. Since
80


there are two distinct activational mechanisms assumed to be
involved in the regulation of this kinase, it would be interesting
to explore other potential interactions by members of the MAPK-
like kinases. As stated earlier, the motifs in the C-terminus of
p70 can be recognized by MAPK-like enzymes. In our lab, there
is preliminary evidence that MEKK3, an isoform of MEKK1,
upregulates p70S6K phosphorylation ( Unpublished observations).
MEKK2 and MEKK3 are newly identified kinases similiar in
structure to each other but different than MEKK1, that are
proposed to be involved in the p42 JUN kinase and the p38 MAP
kinase pathway, but their function and proper placement in the
signaling cascade remains to be elucidated. If there is a link
between MEKK3 and C-terminal phosphorylation of p70S6K, this
may add another level of regulation for this kinase. Whether or
not this means anything in the context of regulation of growth
and proliferation remains to be investigated.
81


APPENDIX
Equipment
DNA Thermal Cycler:
Perkin Elmer Cetus, Inc
Norwalk, CT
Water Jacketted Incubators:
Forma Scientific a division of Mallinckrodt Inc.
Marietta, OH
Mammalian cells 37C 5%C02
Insect cells 30C 5%C02
Laminar Flow Culture Hood:
Biogard Hood
Baker Company
Sanford, Maine
Scintillation Counter:
Beckman LS 5000 TA
Beckman Scientific
Fullerton, CA
Spectrophotometer:
Beckman DU-65
Beckman Scientific
Fullerton, CA
Cell Counter and Channelyser:
Coulter Counter ZM
Coulter Scientific
Miami, FL
Power Supply:
Biorad Model 3000Xi
82


BioRad, Hercules, CA
Centrifuges:
Large Volume;
GPKR centifuge, Swinging bucket 8000 RPM
Beckman, Fullerton, CA
Small volume;
Marthon 13K/M 12,000 RPM
Fisher Scientific
Pittsburgh, PA
Speed Vacuum:
Speed vac plus SC110A
Savant, Farmingdale, NY
Gel dryer:
Drygel Sr., Slab Gel Dryer
Model 1160
Hoefer, Scientific
San Fransisco, CA
Film and Cassettes:
New England Nuclear Life Sci. Products
Reflection Autoradiography film #679062 8X10 In.
Cassettes;
Fisher, Biotech. FBAC 810 Autoradiograghy cassette
Protein Electrophoresis and Blotting:
Mini-Protein II Electrophoresis Cell
1.0 or 1.5 mm thick spacers, 10 well
BioRad, Hercules, CA
Mini-Trans-Blot Electrophoresis Transfer Cell
BioRad, Hercules,CA
Nitrocellulose:
83


Hybond, ECL
Amersham, Life Sciences
Arlington Heights, IL
Two-Dimensional electrophoresis:
HTLE-7000
Hunter Thin-layer Electrophoresis Syste
C.B.S. Scientific, Del Mar, CA
Cellulose Plates:
DC-Fertigplates, Cellulose, EM-5716-7
VWR, Denver, CO
CHEMICALS
Fisher Scientific
Methanol A408-4
Pyridine BP-1155-500
Hydrogen peroxide H323-500
2-Propanol A451-1
Ethyl Ether E138-1
Acetic acid A38c-212
Formoc acid A119-1
Potassium phosphate dibasic 3252-1
Potassium phosphate monobasic 3246-01
Magnesium chloride M33-500
Siama. ST. Louis. MO
Phosphoric acid P6560
2-Mecapto Ethanol M6250
Laurel Sulphate L4390
Sodium Chloride S9888
Citric Acid C8532
Trizma Base T8524
Ammonium Bicarbonate A6141
84


EGTA E4378
EDTA EDS
Beta Glycerol phosphate G6376
Tricine T7911
Triton X-100 T8678
Trichloro acetic acid T-8657
Sodium orthovanadate S6508
Potassium chloride P4504
Isobutyric acid 11754
Alpha Chymotrypsin C3142
dTT (dithiothreital)
Worthington Biochemical Corp.. Freehold. NJ
Trypsin- TPCK treated
Boehrinaer Mannheim
Protein gel mix 168582
FMC
Seakem agarose 50074
Promeaa Biochemical. Madison WS
Neuramindase 480717
Gemini
L-Glutamine 400-110
Pen/strep
Calbiochem
Deoxycholic acid 264101
NP-40 492017
85


Pharmacia
Ficoll-Paque 17-0840-03
BioRad
Ammonium Persulphate 161-0700
Gibco-BRL. Gaithersburg. MD
SF9-900 II SFM media 10902-096
Temed 15524-010
RPMI 1640 media 21870-076
Dulbeccos Modified Eagle media 11855-084
Sodium phosphate 11971-028
Opti-MEM 11057-021
Hepes Buffer 15630-080
Summit Biotech. Fort Collins. CO
Serum- Fetal Bovine Serum
Radioactivity
NEN Dupont. Boston. MA
Easytide BLU-NEG-502X
gama 32P ATP
1 Omci/ml
Phosphorous 32 inorganic
NEX-053
86


CELLS
Human Perpheral Mononuclear Cells
Childerens Hospital
Blood donor center/ Aphoresis unit
Denver, CO
Cos-7 cells
Kidney, SV-40 transformed, African Green Monkey cells
ATCC (American type Culture Collection)
Cold Spring Harbor, NY
Insect cells
SF9 cells (spodoptera frusipeda)
fall armyworm larval cells
Gibco-BRL, Gaithersburg, MD
Sheep Red Blood cells
Colorado Serum CO
Denver, CO
REAGENTS
Talon Metal Affinity Resin Clontech 8901
Nickel NTA resin YO 2281, Quiagen
Cell fectin reagent, Gibco-BRL 10362-010
Xenopus CDK1/cyclin B and CDK2/cyclin E
Jim Mailer
University of Colorado Health Sciences Center
Denver, CO
Histone H1 protein, Boehringer Mannheim, CA 1004875
Rainbow Marker, RPN 756
Amersham, Arlington Heights, IL
87


Bac to Bac Baculovirus expression system
Gibco-BRL,
pfast HT expression vectors 10584-019
rec-Protein G Sepharose 4B 10-1242
Zymed, San Fransisco, CA
SOLUTIONS
Ripa buffer
50 mM Tris (pH 7.5)
150 mM NaCI
1% NP-40
0.5% sodium deoxycholate
0.1% SDS
p70S6K Ivsis buffer
10 mM KP04
1 mM EDTA
5 mM EGTA
10 mM MgCI2
50 mM beta Glycerol phosphate
1mM sodium vanadate
2mM dTT
40 ug/ml PMSF
0.1% NP-40
0.5% Aprotinin
0.5% Leupeptin
p70S6K kinase buffer
20 mM MgCI2
40mM Tris (pH7.5)
2 mg/ml IP-20
0.2 mg/ml BSA
0.8 mM dTT
88


Full Text

PAGE 1

DIFFERENTIAL ACTIVITIES OF CYCLIN-DEPENDENT KINASES IN THE PHOSPHORYLATION OF THE C-TERMINUS OF P70 S6 KINASE. by Philip J. Papst B.A., University of Colorado, 1989 A thesis submitted to the University of Colorado at Denver in partial fulfi IIment of the requirements for the degree of Master of Arts Biology 1997

PAGE 2

This thesis for the Master of Arts degree by Philip Jules Papst has been approved by Naohiro Terada Date

PAGE 3

Papst, Philip, J. Biology) Differential Activities of Cyclin-Dependent Kinases in the Phosphorylation of the C-Terminus of P70 86 Kinase. Thesis directed by Assistant Professor Naohiro Terada Abstract p70 86 Kinase is the major physiological kinase that phosphorylates 86 protein of the 40 8 subunit of eukaryotic ribosomes. p70 86 kinase (p70S6K) has a set of phosphorylation sites located in the carboxy-terminus at positions 8er 411, 418, 424 and Thr 421 that are phosphorylated by an unknown mechanism. These sites represent a recognition motif found in substrates of cyclin-dependent kinases (CDKs). It had been noted previously that cell cycle regulatory protein CDK1 can phosphorylate the carboxy-terminal sites in vitro. Here investigated whether this activity is restricted to CDK1 or more general to other cell cycle regulatory proteins like CDK2 since both CDK1 and CDK2 share a high level of amino acid similiarity and similiar phosphorylation activity toward Histone H1. Using recombinant baculovirus technology, I have generated wild-type and mutant p70 86 Kinase enzymes that provide suitable subtrates to investigate differential phosphorylation events. ill

PAGE 4

demonstrate that a Xenopus CDK1/cyciin B complex but not a CDK2/cyciin E complex phosphorylates all the sites at the C terminus of p70 S6 kinase, especially at serine 411 in the KIRSPRR sequence. The CDK1/cyciin B complex phosphorylated the KIRSPRR peptide with an apparent Km of 50 IlM; whereas CDK2/cyciin E complex did not phosphorylate the peptide well (Km > 500 IlM). Additionally, CDK1/cylin Bcomplexes immunoprecipitated from the Iysates of proliferating human T lymphocytes were able to phosphorylate the KIRSPRR peptide very well, however, immunoprecipitated CDK2, 4, or 6 complexes showed poor phosphorylation activity toward the peptide. p13suc-immune complexes which associate with both CDK2 and CDK1 phosphorylate histone H1 peptide (KKSPKK) at about the time when CDK2 is activated in cell cycle progression of primary T lymphocytes (the G1/S phase of the cell cycle). By contrast phosphorylation of the KIRSPRR peptide of p70 S6 kinase is increased later in parallel to CDK1 activation, at the G2/M boundary. Taken together, these data demonstrate differential specificity of CDK1 and CDK2 on a specific peptide sequence of p70S6K, and support the idea that CDK1 may playa role in the regulation of p70S6K. lV

PAGE 5

This abstract accurately represents the content of the candidate's tresis. I recommend its publication. Signed Naohiro Terada v

PAGE 6

DEDICATION would like to dedicate this thesis to my family. wish to dedicate this thesis to my wife, Marylyn. I wish to thank her for her love and continual support. I will be forever grateful for her kind words of encouragement and inspiration. My education has taken up a large part of our lives, and thanks Marylyn for being there every step of the way. Thanks also to my children, Colton' and Mallory, for their love and joy they show me every day. I wish to thank my parents Linda and Junior Manning for their never ending support and especially for the financial support in my last couple semesters. I could not have completed the degree program withqut all of your help, love, and encourgement. I also would like to thank my father and mother-in-law, Carolyn and Wes Wilklow for all of their general support and especially for their financial assistance towards the end of my degree program. Thank you for always be ing there when I needed someone to talk to. Finally, I would like to thank Kelly and Holly Miller for the donation of their computer. They made it possible for me to spend countless hours working with this computer at home, and since I was not in a position to purchase a computer at the time, their thoughtfulness made a huge difference during the writing of this manuscript. Thank you so very much, I could not have done it without your help.

PAGE 7

ACKNOWLEDGMENTS would like to thank Dr. Naohiro Terada for all his support and guidance with my thesis project these past few years. He has taught me the skills and techniques to be a good scientist. I am forever grateful for your patience and friendship you have shown me throughout the pursuit of my Master's degree. One of the best ways to expand your intellect, is to learn by example. Dr. Terada has shown me the best example of an impressive, extremely bright, and hard working scientist who knows how to get the job done. At the same time exhibiting the compassion and understanding needed in the training of a graduate student. Thank you, Dr. Terada, for the opportunity to work for you and thank you for being such a great mentor. I would like to thank Dr. Linda Dixon for her help with University guidelines and policies. She has been an excellent academic advisor and has guided me through my long and challenging Master's degree program with patience and compassion. Thank you Dr. Dixon for going that extra mile for me and keeping me on track with all the requirements necessary to complete this endeavor. I would also like to thank Dr. Martin Klotz for his critical analysis and helpful commentary of my thesis manuscript. I am also very grateful for his critique of my thesis defense presentation and organization. Thanks to your helpful comments, I have learned an enormous lesson about scientific communication. I would also like to acknowledge Dr. Masayuki Nagasawa and Dr. Hirotaka Sugiyama for their technical help and assistance in the laboratory. Dr. Sugiyama helped with the initial construction and cloning of the wild-type mutant p70 S6 Kinase proteins.

PAGE 8

Dr. Nagasawa provided assistance for many of the experiments presented here and especially with the phosphorylation assays. Finally, I would like to thank Jacque Horvath for her computer knowledge and technical writing assistance during preparation of the manuscript. Thanks to her I learned a fair amount about Microsoft Word and all of its capabilities. Thank you Jacque for always taking the time to answer all my questions.

PAGE 9

CONTENTS CHAPTER 1. INTRODUCTION .................................. ............... 1 Background........................................................................ 1 Protein Serine/Threonine Kinases ....... .................. 2 The Cell Division Cycle....................... ............................ 3 The History of p70 S6 Kinase.................... ................. 7 p70 S6 Kinase and Translation....... ..... ........ .......... 1 1 The Use of Rapamycin as a Tool to Investigate the Role of p70 S6 Kinase in Cell Proliferation... 1 4 Signaling Pathways... ...... .... ......... .... ......... .......... ........... 1 5 The p70 S6 Kinase Molecule: the Domains of the Kinase and Their Putative Functions.............. 1 7 Model of p70 S6 Kinase Regulation. .... .... .... ... ... ... ... 20 The CDK Substrate Recognition Domain............. ... 23 The Discrepancy................... ............... ....................... .... 2 4 Hypothesis .......... ......... ............................................... ... 2 4 ix

PAGE 10

2. MATERIALS AND METHODS............................................. 26 Cell Culture......................................................................... 2 6 Antibodies and Immunoprecipitations.................. 27 Baculovirus Expression System. ....... .... ..................... 2 8 Construction of the p70 S6 Kinase Clones............. 29 Transformation and Transposition of BAC p70 Clones into BAC 10 Cells for Integration of p70 S6 Kinase Into the Baculovirus Genome....... ......... 3 1 Isolation and Analysis of Recombinant p70 S6 Kinase Bacmid DNA for Transfection and Viral Production....... ... ..... ......... ........ .... ......... ...... ...... ....... .... ..... 3 2 Transfections...................................................................... 33 Infections....... ..................................................................... 3 4 CDK1 /Cyclin B Phosphorylation of Recombinant p70 S6 Kinase.... ................................... 3 5 Phosphopeptide Mapping......................... .... ............... 3 6 Protein Precipitation and Digestion...... ....... 3 6 Electrophoresis/Chromatography................ 37 Synthetic p70 S6 Kinase Ser 411 Phospho-Peptide Map................... ....... ............. 3 8 CDKlCyclin In Vitro Kinase Assay............................ 39 x

PAGE 11

T Cell Time Course..... ................................................. ... 4 0 T Cell CT -p70 S6 Kinase In Vivo Phosphorylation.... ................. ..... .... .... .... ............ 4 0 Activity Time Course of p70 S6 Kinase and CDKl........ ............................. .......................... 41 Cell Cycle FACS Analysis. ....... ........... ............ ... 4 2 Increase of CT -p70 S6 Kinase Phosphorylation by Nocordizole Treatment In Vivo..... ... ........ .... ...... 42 3. RESULTS...................... ............................................................. 44 Baculovirus Expression System Provides High Quantities of Functionally Active Wild-Type or Mutant p70 S6 Kinase for Phosphorylation Analysis... ... ......... ....... ... ... .... ....... ....... ......... ...... ..... ......... 4 5 CDKlICyclin B and Not CDK2/Cyclin E Phosphorylates Recombinant CT -p70 S6 Kinase on Ser 411.... .............. ............ ..... ................... 4 8 The Rate of Peptide Phosphorylation by CDKlICyclin B Complex is Higher Than CDK2/Cyclin E Complex............................ .................. 57 CDKlICyclin B Preferentially Phosphorylates CT -p70 S6 Kinase Whereas Other Members of the CyclinDependent Kinase Family Do NOL.... 61 Differential CDK 1 Activity on CT-p70 S6 Kinase CTerminal Peptide at a Different Cell Cycle Phase Than Histone HI Peptide .................. 6 3 xi

PAGE 12

Late Cell Cycle Phosphorylation Versus Early Activity of p70 S6 Kinase....... ........ ......... ............... ..... 6 5 The Enhancement of CDKI Protein Expression and Activity With Nocordizole Treatment Also Markedly Increases p70 S6 Kinase Phosphorylation In Vivo.............................................. 7 2 4. DISCUSSION............................................................................. 7 6 Implications for CDKI Phosphorylating p70 S6 Kinase................................................................................... 77 Subcellular Localization................................................ 7 8 Nucleolar Organization............... .................................. 7 9 Future Experiments ................................................. ...... 8 0 APPENDIX....................................................................................... 8 2 Equipment. ......... ...... ......... ... ..... ... .......... ............. ..... ........ 8 2 Chemicals... ...... .... ...... ..... ... ... ........ ... ...... ...... ....... ... ...... ... .... 8 4 Cells....................................................................................... 87 Reagents.... ...... ........... ...... .................. ........... ... ....... ..... ... .... 8 7 Solutions.............................................................................. 8 8 REFERENCES.................................................................................... 90 xii

PAGE 13

CHAPTER 1 Introduction Background One of the most fundamentally important aspects of biology is how cells grow and proliferate and how signals generated at the plasma membrane by various stimuli are transmitted to the nucleus, whereby the appropriate response can be made. The events leading up to and encompassing cell division are enormously complex (Patel et.a!. 1991). To understand how cells communicate, it is essential to first isolate individual molecular interactions that make up a particular response and then analyze it in the context of other reactions occuring as a result of the stimulus. Cell surface receptor ligation induces a cascade of phosphorylations and dephosphorylations transduced by kinases and phosphatases. Many of these enzymes function in series as part of a specific pathway resulting in increased transcriptional and translational activity required for proliferation. Several of these pathways function in parallel generating a complex network of signal transductions all induced by the same growth stimulus. 1

PAGE 14

One of the goals of molecular biology today is to dissect these pathways to understand how cells communicate at the molecular level. The molecular enzymology of the protein kinases mediating these communications will provide the necessary framework upon which to acheive these goals. Protein Serine-Threonine Kinases There are two major eukaryotic families of protein kinases, the tyrosine kinases and the serine/threonine kinases (Hanks et al. 1988). These kinases have the ability to add phosphorous to specific tyrosine, serine or threonine amino acids of other proteins thereby activating them to carry out their specific functions. The activity of serine-threonine kinases play an important role in eukaryotic cell physiology including control of cell proliferation. Unlike the tyrosine kinases which are more implicated with phosphorylation events coupled to cell surface receptor, serine/threonine kinases function intracellularly, phosphorylating numerous proteins in series that function as downstream effectors of signal transduction pathways. There are several members of this family including protein kinase C, the cAMP-dependent protein kinases, the cyclin-dependent kinases controlling the cell cycle, and the Raf signalling complex. Raf is an excellent example of a serine-threonine kinase that is activated downstream of the G protein Ras and then serving to 2

PAGE 15

activate a protein kinase cascade involving the MAP kinase pathway that ultimately triggers cell proliferation. Protein serine-threonine kinases are more abundant in the cell than their tyrosine kinase counterparts Amoung these serine-threonine kinases is the enzyme called p70 86 Kinase (p70S6K), which lies in a signal transduction pathway distinct from the MAP kinases (Ming et al. 1994). As will be discussed in more detail later, p70S6K is a key mediator in the regulation of ribosome biogenesis and the subsequent regulation of protein synthesis that is required for cell proliferation. The Cell Division Cycle Cell growth is divided into distinct phases, each representing different functions as the cell prepares to go through mitosis. The cell cycle is made up of G1 (gap 1), a preparatory phase before S phase (DNA sythesis), and G2 (gap 2), which prepares the cell for mitosis (M). Growth factors stimulate cells to progress through this cycle, which is controlled by a family of protein serine-threonine kinases that are highly conserved through evolution. These kinases are called cyclin dependent kinases (CDKs) because they associate with proteins referred to as cyclins (Figure 1). CDKs are thought of as the active component of the CDKlcyclin complex. Together, they along with a host of other regulatory molecules help guide the 3

PAGE 16

entry into, progression of, and exit from the cell division cycle (Pines J. 1993). More importantly, they function at two main controlling points, of the cycle. They control the progression of cells from G1 to S and from G2 to M. When quiesent cells receive a signal from a growth factor or other mitogen, different distinct combinations of CDKs and their associated cyclins assemble into complexes that provide the necessary kinase activity to traverse through the control points and progress through the cell cycle. 4

PAGE 17

CDK1/ cyclin B CDK2/ cyclin A CDK2/ cyclin E Go CDK4,6/ cyclin 0 Figure 1. Differential roles of cyclin-dependent kinases and their as ociated cyclins in cell cycle regulation. 5

PAGE 18

The two main CDK's studied in this project, are CDK1 (also referred to as p34cdc2) and p33CDK2. CDK1 associates with either cyclin A or cyclin B and is active in G2 and M phases of the cycle. The G1/S transition is controlled by CDK2 and its cyclins, either cyclin A or cyclin E (Elledge et al. 1991). The initiation of the growth cycle is regulated by the 0 type cyclins and its association with CDK4 and CDK6 (Pine J. 1993; Sherr C.J. 1993). In each case, the components of the complex physically interact with each other, subsequently becoming active specifically phosphorylating various substrates that coordinate numerous transcriptional activities that prepares the cell for division. Then the complex dissociates in time for the next family of CDKlcyciin complexes to carry out their function. Substrate specificity is an important attribute of each CDK/cyclin complex. CDK1 is responsible for the breakdown of the cytoskeleton and the nuclear envelope by phosphorylating many of their structural components. CDK1 also phosphorylates Histone H1 thereby regulating chromatin structural components in M phase (Migg E. 1993). Alternatively, CDK2 and its associated cyclins phosphorylate a protein complex called Rb ( the retinoblastoma protein), which is known to be an inhibitor of the transcription factor E2F. The phosphorylation of Rb releases E2F and allows it to bind to specific promoter regions of DNA in the nucleus thereby initiating new gene expression needed for cell division (Nigg E. 6

PAGE 19

1993) Given the importance of CDK1 and CDK2 in regulating the G2/M and G1/8 transitions, the search for other substrates of these serine/threonine kinases might lead to new pathways of regulation important for cell proliferation. The History of p70 86 Kinase In the late 1980's, due to the role of protein phosphorylation cascades involved in the activation of cell growth and the importance of ribosomal 86 protein for the regulation of the mRNA translation and the control of protein sysnthesis, it was imperative to identify the kinases controling these processes. A protein was found in cell extracts with potent 86 activity that itself was sensitive to phosphatase treatment indicating that this protein itself was phosphorylated (Jeno et al. 1988). The putative 86 activity was isolated, purified to homogeneity, and identified as a kinase with the apparent molecular weight of 70 kilodaltons. Upon cloning, two transcripts were identified, differing only in 23 amino acids at the amino terminus (Grove et al. 1991). This other transcript produced by differential mRNA splicing, was called p85s6K ( Banerjee et al. 1990). It contains a cluster of basic amino acids (Arg Arg-Arg-Arg-Arg-Arg) that comprises a nuclear translocation motif in these 23 amino acids. Nuclear localization of p85s6K was shown by using microinjected expression vectors 7

PAGE 20

containing p85s6K or fusion proteins containing the first 150 amino acids of p85s6K (Reinhard et al. 1994). It has been estimated that approximately 4000-6000 p70S6K molecules are expressed in a single cell, as compared to only about 300 S6K molecules of p8S (Kozma et aI., 1993). It was proposed that this form of the enzyme maintains a pool of phosphorylated 86 protein in the nucleus but its actual function is unknown (Franco and Rosenfield, 1990). The p70 constituent is active primarily in the cytoplasm and is now known to be the major physiological kinase that phosphorylates 86 in the cleft of the assembled ribosome where the mRNA is bound and protein synthesis is initiated (Ballou et al. 1991, Blenis et al. 1991). There are actually two families of 86 kinases: the p90rsk encoded kinases and the p70 and p8S 86 kinases. The, p90rsk family of kinases only phosphorylate 86 kinase in vitro. The p70/85 86 kinases are the major physiological kinases that phosphorylate 86 in vitro and in vivo. There is considerable evidence that places the two different kinases on distinct signalling pathways. Unlike p70S6K p90rsk is phosphorylated and activated by the p42 and p44 mitogen-activated protein kinases (MAPK) in response to signals transmitted from the RA8/RAF/MEKK series of kinases that are activated by growth factor receptor stimulation (Ming et al. 1994; Blenis et al. 1991). 8

PAGE 21

The p90rsk and MAP Kinases then trans locate to the nucleus to phosphorylate transcription factors, thereby promoting transcription of genes required for proliferation. The p70 86 kinases are completely distinct from this signalling pathway and have a function all their own. The p70S6K is known to be in a signalling pathway that places it below the lipid kinase PI3K (Figure 2). 9

PAGE 22

PTK Figure 2. Overview of growth factor mediated signal transduction pathways leading to gene expression or protein synthesis. p70 S6 Kinase lies in a distinct pathway from the MAP kinases. 10

PAGE 23

p7086 Kinase is a serine/threonine intracellular kinase that is inactivated in quiescent cells, and upon activation has been shown to phosphorylate the ribosomal 86 protein in higher eukaryotes (Ballou et al. 1991; Blenis et al. 1991). This enzyme is an intermediary in a signal transduction cascade of phosphorylations initiated at the plasma membrane by many known mitogenic stimuli including EGF, PDGF, IL-2, IL-3, GM-C8F, and insulin. This kinase has been shown to be important for quiescent cells to enter the cell division cycle (GO/G1), as indicated by the inhibition of p70S6K by the specific inhibitor rapamycin or a micro-injected antibody against p70S6K (Lane et al. 1993; Chung et al. 1992). These inhibitors either arrest the cell in G1 or cause G1 prolongation when cells have already entered the cell cycle (Terada et al. 1993). P7086 Kinase specifically phosphorylates multiple sites on ribosomal 86, a protein of the 408 ribosomal subunit, and thus is coordinately involved in the regulation of protein synthesis (Bandi et al. 1993). P70 86 Kinase and Translation p70S6K phosphorylates ribosomal protein 86 (Ballou et aI., 1991, Blenis et aI., 1991). The ribosomal 86 protein is located in a position in the 40 8 subunit that is proposed to be the major functional site involving mRNA and tRNA binding (Bommer et aI., 11

PAGE 24

0; Terao and Ogata,1979; Tolan and Traut, 1981; Geyer et al. 19 2; Kjeldgaard et al. 1994). After mitogenic stimulation, S6K phosphorylates several sites in the carboxy-terminus of a region proposed to lie in the cleft where mRNA binds, su gesting a role in the regulation of translation (Kozma et al. 1989). Nielson et al. suggested that 86 phosphorylation may be involved in the recruitment of mRNA's into polysomes because an ibodies to 86 block the binding of elf-2/GTP/MET-tRNA plexes to 40 8 ribosomes. This suggested that phosphorylated may be directly involved in translation initiation and directly m dulate mRNA binding in ribosomes during protein systhesis. s increasing p70S6K activity increases S6 phosphorylations, su sequently increases translational activity, especially peptide ch in initiation. The correlation to translational regulation is not associated overall protein synthesis, because there is a discrepancy be ween 86 phosphorylation and the overall rate of protein sy thesis( Thomas et aI., 1980,1982). Many cell lines like Swiss 3T or Cos cells when stimulated with EGF or other growth fa ors increase 86 phosphorylation rapidly (within 30-60 utes), and then as cell cycle progresses the phosphorylation of S6 gradually decreases. Overall protein systhesis however, inc eases more slowly reaching a maximal point after 3-4 hours. Th refore, it was suggested that 86 phosphorylation is fu ctioning in a mechanism of selective translational control 12

PAGE 25

instead of regulating overall protein synthesis( Thomas et al. 1992}. Subsequently, this established the current dogma that p70S6K regulates translation of specific mRNA's through the action of S6 phosphorylation. Taking this notion to a higher level, a mechanism was proposed for this kind of regulation. All ribosomal proteins in higher eukaryotes have a consensus sequence at their transcription initiation site (Jefferies et al. 1994) This sequence typically consists of a span of pyrimidines that follows a cysteine in the transcription initiation site designated CYYYYYY. This sequence is followed by a GC rich region. Interestingly, in mammalian cells, both elongation factor 1 a (eEF-1 a) and elongation factor 2 (eEF-2) have this same consensus sequence at the 5'-terminus of their mRNA, but not in their genomes (Nakanishi et aI., 1988; Uetsuki et aI., 1989). It is possible then that this sequence is correlated to selective translational regulation and indeed, several groups have since shown that this 5' polypyrimidine tract is required for the translational control of r-bosomal protein mRNA's and elongation factor mRNA's (Levy et al.). Thus the activation of p70S6K leads to the phosphorylation of S6 which inturn upregulates the translation of selective mRNA's encoding the components of the ribosome. The ribosomal proteins and elongation factors need to be present before general protein systhesis can begin. This then might possibly explain the discrepancy between early and rapid 13

PAGE 26

S6 phosphorylation and later increases in overall protein systhesis. The Use of Rapamycin as a Tool to Investigate the Role of p70 S6 Kinase in Cell Proliferation Much of the knowledge obtained to date about the regulation of p70S6K has been deduced from experiments using the antifungal antibiotic Rapamycin (RAP). Rapamycin is a macrolide with functional similarities to cyclosporin and FK506; they are all potent immunosuppresants and have been shown to prolong survival of organ allografts and prevent autoimmune diseases (Morris R. E.). RAP inhibits proliferation of lymphocytes as well as other types of cells by arresting cell cycle progression at early G1 phase or delaying entry into S phase if mitogen stimulated quiesent cells have already entered the cell cycle (Terada et aI., 1993). RAP binds to FKBP12 (FK506 binding protein) and this associates with a third molecule called FRAP or RAFT, and it is this complex that is responsible for the indirect inhibition of p70S6K ( Brown et aI., 1994,1995; Weng et aL, 1995; Schreiber et aL 1991}. This FRAP protein has considerable homology with the catalytic subunit of PI-3 kinase, and has since been demonstrated by Brown et al to be located upstream of S6K p70 Our lab and others have demonstrated that the RAP/FRAP complex specifically dephosphorylates and inhibits the activity 14

PAGE 27

of p70S6K at Thr 229, Thr 389, and Ser404, with no alterations in many other early signal transduction events after mitogen stimulation ( Chung et aI., 1992; Kuo et aL, 1992; Price et aI., 1992; Calvo et aL, 1992; Terada et aL, 1992). Indeed, through the use of recombinant DNA techniques the substitution of Thr 229 to alanine eliminates the sensitivity of p70S6K to RAP and makes an inactive kinase (Sugiyama et aI., 1996). The Thr 389 phosphorylation site has also been shown to be a Rapamycin sensitive site (Thomas et aI., 1995). Indeed, the motif surrounding the Thr 389 is highly conserved in many members of the second messenger family of protein kinases including the PKC isoforms. This domain has been implicated as a crucial sequence in the maintenance of kinase activity in these other kinases. Rap has no effect on the other phosphorylation sites in the carboxy-terminal region (Ferrari et al. 1993; Pearson et al. 1995 ). Signalling Pathways The current signalling pathway of p70S6K and other associated pathways form a complex network of signal transduction events during mitogenic or growth factor stimulation. One of the signalling pathways activated after cell 15

PAGE 28

surface receptors are ligated includes the p70S6K pathway. The upstream kinases or phosphatases remain to be clearly identified, however there is preliminary evidence indicating phosphatidyl inositol-3 kinase (PI-3 kinase) and protein kinase B are located upstream of p70S6K (80S J.L.,1992 and 80udewijn et aL, 1992; Burgering et aL 1995). PI-3 Kinase is one of many lipid kinases that are part of a pool of activated second messenger molecules stimulated after EGF and PDGF receptor ligation. PI-3 kinase phosphorylates phosphoinositol 3,4,5-triphosphate which then through a series of proposed intermediaries, leads to p70S6K activation (Downward. J., 1995). Neither of these kinases directly phosphorylate p70S6K either, demonstating other as yet unidentified kinases involved in phosphorylating p70S6K Other mitogen-stimulated signalling pathways bifurcates at the level of the p21 RAS, thus making the ERK-encoded MAP kinase 1 and 2, the p38 MAP kinases, the p42 jun kinase and pathways transmitted from protein kinase C (PKC) all distinct signalling cascades from the p70S6K signalling pathway (Ming et aL, 1994). 16

PAGE 29

The p70 S6 Kinase Molecule: the Domains of the Kinase and Their Putative Functions The p70S6K protein is 502 amino acids long and consists of several domains important for its function (Banerjee et aI., 1990). These domains were characterized with the help of rapamycin studies. The cataylic domain comprising residues 89355 contains the threonine 229 and Thr 389 sites that are the principal phosphorylated sites responsible for kinase activity; the loss of these sites renders the kinase inactive (Sugiyama et aI., 1996; Pearson et aI., 1995). The catalytic domain also exhibits an ATP-binding site (GKGGYG) located at amino acid 98103 functionally similiar to other known kinases. It has been postulated that there are two independent pathways which regulate kinase activity, group 1, located at positions Thr 229, Thr 389, and Ser 404 is responsible for its activation, and the putative translational regulation of specific ribosomal protein and elongation factor mRNA's (Han et aL, 1995; Pearson et aL, 1995). This phosphorylation is inhibited by the immunosuppressive drug Rapamycin, but the T229 and T389 phosphorylation sites are differentially regulated by different kinase kinases that are Rapamycin insensitive (Dennis et al. 1996). Group 2 is located in the carboxy-terminus at positions Ser 411,418,424 and Thr 421 (Ferrari et aI., 1993; Weng et aL, 1995). This site is rapamycin insensitive. Numerous reports 17

PAGE 30

have focused on group 1 phosphorylation site and its involvement in translational regulation, however, little is known about group 2 phosphorylation activity and its physiological consequences. Phosphorylation of group 1 occurs early in the cell cycle whereas the C-terminal site phosphorylation occurs in later stages of the cell cycle ( Lane et al. 1993; Susa et al. 1989; Weng et al. 1995). 18

PAGE 31

rP70S6k rp85s6k acidic region :tiva ion of the kinase Fig re 3. Protein structure MAPK, CDK-like kinase ??? '" ... Set 2 eOOH pseudo-substrate site Set 1 rapamycin, wortmannin-sensitive Pl3K-sensitive 19

PAGE 32

Model of p70 86 Kinase Regulation The culmination of many years of investigation into the regulation of this kinase has produced a model of that attempts to explain many of the findings(8anerjee et al. 1993). Residing in the catalytic domain is the Thr 229 phosphorylation site that is essential for kinase activity and also contains the ATP binding site. This site is proposed to be blocked by a portion of the carboxy-terminus that contains a putative pseudosubtate sequence. This basic pseudosubstrate site is envisioned to bind to and occlude the substrate binding site including the ATP binding domain and maintain the kinase in a basal-inhibited state. This carboxy-terminal region is refered to as an autoinhibitory domain. There is an acidic region (Asp29-Glu46) slightly amino terminal of the catalytic domain that is thought to be a candidate for the complementary substrate binding site; a region that could provide a suitable binding site for the highly basic protein ribosomal 86. Thus the basic pseudosubstrate region in the carboxy-terminal tail is thought to fold over and inhibit the cataytic domain by binding to this upstream acid region when the kinase is in the inactive form. Upon activation of quiescent cells, p70S6K phosphorylations in and around the C terminal pseudosubstate region weakens the electrostatic binding of the two complementary regions thereby releasing the occluded 20

PAGE 33

catalytic domain, permiting access to ATP and 86 and allowing the enzyme to be activated via the group 1 phosphorylation sites. The Thr 389 site of group one actually lies in the linker region and its phosphorylation would also help in this conformational change that exposes the catalytic site (Pearson et aI., 1995). 21

PAGE 34

Os6k Thr389 Figure 4. Proposed model for the regulation of p70 S6K 22

PAGE 35

The CDK Substate Recognition Domain Each of the four carboxy-terminal phosphorylation sites comprise the sequence Ser/Thr-Pro motif which is often recognized as a substrate for CDk-like and MAPK -like proteins ( Banerjee et aI., 1990; Kitagawa et al. 1996). The specific sequence RSPRR of the ser 411 region is the actual consensus sequence of CDK1 substrate recognition motif ( Songyang et aI., 1994). I speculated that this sequence present in the carboxy terminus of p70S6K had some importance in the regulation of the enzyme, and thus further explorations into the functional significance that this might imply might lead to novel mlecular interactions crucial to the control of proliferation. Given the evidence to date demonstrating two distinct activation domains, there must be other kinases active in the cell that phosphorylate p70S6K and regulate its activity for other important processes in cell physiology. Scanning the literature on p70S6K produced one piece of evidence that CDK1 can indeed phosphorylate these C terminal sequences in vitro ( Mukhopahdya et aI., 1992). This paper only showed CDK1 and did not investigate other cell cycle regulatory proteins. Here I investigate the differential specificities of the major players of the cell cycle regulatory proteins in the regulation of p70S6K 23

PAGE 36

The Discrepancy Between the Timing of the Cell Cycle and the Activation of p70 86 Kinase There was an another intriguing observation made in our laboratory concerning the time course of phosphorylation of p70 in vivo. Many experiments have shown massive phosphorylation of p70S6K late in the cell division cycle of lymphocytes. In contrast, p70S6K activity in these cells has been shown to be very rapid after the cells were stimulated to enter the cycle. If the principal function of the kinase occurs very early and is associated with G1 progression as the previous studies have indicated, why does this enzyme need to be so intensively phosphorylated late in the cell division cycle? I propose that the answer to this question may be found closely correlated with the activity of CDK1 and the control of the G2/M transition. Hypothesis The purpose of this work is to investigate the carboxy terminal phosphorylations of p70 86 kinase and its potential role in late cell cycle regulation through the interaction of cyclin-24

PAGE 37

dependent kinases and there associated cyclins. This project presents and supports the hypothesis that CDK1/cyclin B can phosphorylate a specific serine residue in the carboxy terminus of p70 86 kinase only late in G2/M phase of the cell division cycle. CDK 1 is the principal CDK invovled in this regulation. Both in vitro and in vivo evidence is presented to support this hypothesis and illustrate a direct link of the p70S6K signalling pathway and the cell cycle machinery. 25

PAGE 38

CHAPTER 2 MATERIALS and METHODS Cell Culture Human peripheral blood cells were obtained by leukopheresis of healthy donors from Children's Hospital Blood donor center. Mononuclear cell suspensions were prepared by Ficoll-Hypaque gradient centrifugation and T cells were obtained by Erosette enrichment using nurimidase treated sheep red blood cells (Colorado Serum Co, Denver, CO) as described. Cells were maintained in RPMI 1640 complete medium (Gibco BRL, Grand Island, NY) supplemented with 10% (v/v) heat inactivitated fetal bovine serum (FBS, Hyclone, Logan Utah), 2mM Glutamine (Gibco), 100u/ml penicillin and 100mg/ml streptomycin (Gibco). T cells were stimulated with Phorbol 12,13-dibutyrate (PDBu, Sigma, St. Louis,MO) and calcium ionophore, ionomycin (Calbiochem, San Diego, CA) dissolved in dimethyl sulfoxide (OMSO). Rapamycin was obtained from the Drug Synthesis and Chemistry Branch, NCI, and dissolved in PBS to give a 0.1 mg/ml stock solution. 26

PAGE 39

Insect cells (Spodoptera Frugiperda Sf9) were maintained at 30 degrees in serum-free SF900 \I SFM medium (GIBCO) supplemented with SOu/ml penicillin and SOmg/ml streptomycin. Cells are cultured in a monolayer and become suspended in medium by gentle dispersion. Cells were passed every four days by 1/3 or 1/4 dilution. Antibodies and Immunoprecipitations Antibodies used in this study were; p70S6K a rabbit polyclonal antibody raised against the carboxy-terminus (amino acids S02-S25 NSGPYKKQAFPMISKRPEHLRMNL) anti-CDK1, rabbit polyclonal antibodie raised against human CDK1 (kindly provided by David Beach); anti-cdk2, rabbit polyclonal antibodies raised against an epitope 283-298 of human cdk2 (sc-163, Santa Cruz Biotechnology, Santa Cruz, CA); anti-CDK4, rabbit polyclonal raised against amino acids 1-303 (Santa Cruz); anti-Cdk6, rabbit polyclonal raised against amino acids 306-326 (Santa Cruz); anti-cyclin A, rabbit polyclonal antibodies raised against human cyclin A (kindly provided by Jonathon Pines); anti-cyclin B, mouse monoclonal IgG1 recognizing epitope 1-21 of human cyclin B1 (GNS-1,Pharmingen, San Diego, CA). Immunoprecipitations were done using Protein G sepharose (Zymed Laboratories, South San Francisco, CA)and p13 suc agarose beads (Oncogene Science, 27

PAGE 40

Uniondale, NY) Typically 5-10 Jl.1 of antibody was used per 0.5S6K 1.0ml of lysed cells in lysis buffer ( RIPA buffer or P70 buffer) for 1-2 hours on ice. Prewashed protein G sepharose (30-50Jl.1) was added and rotated at 4 C for 30 minutes-1 hour. Separose beads were centrifuged and washed twice in lysis and then 1 x loading dye was added and the samples boiled for 5 minutes to remove proteins from sepharose. Baculovirus Expression System Baculovirus expression system was employed to generate large quanties of active and inactive wild type and mutant protein which provided excellent substrates for peptide phosphorylation analysis. This is a viral expression method in armyworm larval insect cells. The insect cells are infected with recombinant Autogragha california nuclear polyhedrosis virus (AcNPV) by means of efficient vectors that allow high levels of heterologous gene expression. Baculoviruses have a restricted host range, only infecting specific invertebrate species. Large quanties of soluble functionally active protein of interest that can be selected specifically by a 6-histidine tag sequence located 5' of the cloned mammalian protein. The vectors employed also have a powerful polyhedron promoter, transposon recognition sequences, a rTEV 28

PAGE 41

protease recognition sequence, a large multi-cloning site, and selectable antibiotic markers. Recombinant DNA clones were isolated and used to transform Sac 10 cells that contains the viral genome. These cells contain the 150kb viral genome and helper plasmid that encodes the transposase enzyme. The DNA between the tn7 sequences including the p70 S6K sequence will transpose into the viral genome at a specfic target that is under the control of lac Z. This allows the inducible expression by IPTG and blue gal and the selection of white recombinant clones against a background of non-recombinant blue colonies. White clones are restreaked to ensure purity and the DNA isolated from these possible positive clones. Large molecular weight DNA is analyzed by 0.5% agarose gel. This DNA is then used for the transfection of SF9 insect cells to obtain complete virus particles that are then used to infect SF9 cells on a large scale. Construction of P70 S6 Kinase Clones The full length p70S6K cDNA clone was kindly provided by George Thomas and initially subcloned into pXST mammalian expression vector. This pXST clone was analyzed by restiction endonuclease cleavage and mammalian transfection to determine efficacy of the clone (Sugiyama et al.,1996). The Sal I restriction S6K endonuclease cleaved a 1.6 kb p70 fragment from these pXST clones that represents the wild type used for further subcloning. 29

PAGE 42

I purified the fragment by separation through an agarose gel with subsequent DNA purification with the Quiagen purification system. This fragment was cloned into the Sal I digested, calf intestinal phosphatase-treated (Gibco BRL) Baculovirus expression vector Fast Sac HTb (Gibco BRL). The phosphatase was used to remove 3' and 5' phosphates on the vector to lower background by markedly reducing vector self ligation. Ligation was done with T4 ligase (Promega) according to standard protocol. Then 1-10ng of ligated DNA's were used to transform DH 5 alpha E-coli subcloning cells (Gibco BRL) and recombinant clones screened by mini prep isolation and digestion with Sal I restriction enzyme to determine the presence of p70 insert DNA. Conformational restriction analysis was done with several other restriction enzymes to determine proper oreintation. The T229A and WT deletion 104 clones were do ned in a similiar manner. The initial DNA was subcloned from previously constructed pXST clones ( Sugitama et al ., 1996). The T229A mutant was generated by PCR specific primers that incorporates Sph I and Nsi I sites at the T to A transition and uses the wild type p70 clone as a template. This cDNA clone was digested with SaIl and Bal I to isolate by agarose gel the T229A mutant fragment and subcloned into the HTb baculovirus vector using a Nhe I linker. The vector was treated with Sal I and Xba I enzymes, Xba I being compatible with Nhe I. Ligaton at the Sal land Nhe I/Xba Isite creates the completed clone. The c-terminal deletion 104 clone 30

PAGE 43

was made by direct subcloning from pX8T vector to the fastbac HTb vector by the Sal land Xba I sites. Multi enzyme digestion was done on these clones as well. Once positive clones were identified, large scale maxi-prep (Gibco-BRL) was done to obtain large quantities of DNA for further experimentation. Transformation and Transposition of BAC p70 Clones into BAC10 Cells for Integration of p70 86 Kinase into the Baculovirus Genome The three p70S6K clones were then transposed (site-specific) into the baculovirus genome via the transformation of DH10 Bac competent bmon14272 ( Gibco) E. coli. cells. These cells contain low copy number mini-F replicon, a kanamycin resistance marker, and a lacZ alpha coding region. In the N-terminus of lacZ protein is an attachment site for the bacterial transposon Tn 7 which does not disrupt the reading frame of lacZ. The native Bacmid propogates in the E. coli. Bac 10 cells and complements the lacZ deletion on the chromosome to form colonies that are blue. The recombinant p70S6K bacmids contain the P70 DNA in the mini-Tn7 site which when transposed into the viral genome disrupts the expression of Lac Z alpha peptide making white colonies. For each transposition, 1 ng of DNA was used in 5111 of water and added to 100ml of Bac10 cells on ice for 30 minutes. The mixture was incubated in a 42C water bath for 45 seconds to heat shock and 31

PAGE 44

then back in ice for two minutes. Then 900 I.LI of S.O.C. medium was added and the mixture was incubated for 4 hours at 37 C. Serial dilutions were plated on Luria agar plates containing 50 mg/ml kanamycin, 7 mg/ml gentamycin, 10mg/ml tetracycline, 100mg/ml Blue-gal and 40mg/ml IPTG. Plates were incubated for at least 24 hours at 37 C to observe colonies. Positive white colonies were restreaked and a we" isolated colony was used for liquid culture for the isolation of high molecular weight recombinant viral DNA. Isolation and Analysis of Recombinant p 70 S6 Kinase Bacmid DNA for Transfection and Viral Production Isolated colonies were sterily transfered to 2ml of LB medium containing 50mg/ml kanamycin, 7mg/ml gentamycin, and 10 mg/ml tetracycline and grown at 37 C for 24 hours shaking at 300 rpm. Cells were centrifuged at 14000 x 9 for 1 minute and resuspended in 0.3 ml of solution 1 [15mM Tris-HCI (phS.O), 10mM EDTA, 100mg/ml RNase A] and then 0.3 ml of solution 2 (0.2N NaOH, 1 % SDS). 0.3 ml of 3M potassium acetate was added and samples spun at 1400x 9 for 10 minutes. DNA in supernatant was precipitated in O.S mls of isopropanol for 10 minutes and centrifuged again for 15 minutes, washed with 70% ethanol, and the DNA pellet air dried for 10 minutes. The DNA was resupended in 40J.L1 of TE for further analysis. Five ml of Bacmid DNA was 32

PAGE 45

analyzed on a 0.5% agarose gel containing 0.5 mg/ml of ethidium bromide in TAE buffer. Samples were run at 23 volts for 12 -15 hours to identify the 135 kb viral genome. Samples positive for Bacmid DNA were used for transfection. Transfections Transfections were carried out by seeding 9 x 105 cells in a 6-well plate containing SF-900 SFM media with penicillin/streptomycin at 0.5X final concentration for 2 hours to attach to the well. Then SJlI of bacmid in 100ml SF-90011 SFM per transfectant was mixed with 100 JlI of SF-90011 SFM containing 6 ml of CellFectin reagent (Gibco) for 45 minutes at room temperature. The lipid/DNA complexes (diluted t01 ml) was overlayed onto the insect cells in the 6-well plates seeded earlier, each containing 0.8 mls of SF 90011 SFM. The complexes were allowed to enter the insect cells by incubating for 5 hours at 30 cC. Then 2 ml of SF90011 SFM was added to each plate and incubated for 48 to 72 hours. Protein expression and activity was analyzed after 48 hours and virus particles were harvested after 72 hours. Viral titers were determined by viral plaque assay. Briefly, 6 x 10 5 insect cells were seeded into 6-well plates for 2 hours to make a unifrom monolayer. An 8 log serial dilution was made of the virus harvested from the 72 hour transfections 33

PAGE 46

and 1 ml was layered onto the monolayer on insect cells for 1 hour. A 4% agarose solution was set at 70C and 1.3 X SF90011 insect medium was placed at 40C for making the SF-900 plaquing overlay solution. Mixed 30ml of 1.3X insect media with 10 ml of the agarose solution and placed at 40C. Sequentially, removed viral supernatants form cells and replaced with 2 ml of diluted agrarose/media solution. Agarose was allowed to harden for 20 minutes and the plates were incubated at 30C for 4-14 days. Viral plaques were visible after 5 .. 7 days and at the proper dilution, 3-20 plaques were counted after no changes in plaque number was observed. Typically, viral titers were 1-2 X 10 6 pfu/ml. Viral titers were amplified by sequential infections to achieve 1 X 10 8 pfu/ml. Infections Viral infections were done with 0.5-1'.0 ml of transfection supernatants in 30 mls of insect cells (2 X 10 6 cells/flask) for 48-72 hours. Cells were harvested and lysed in Ripa buffer or p70S6K buffer and the 6 HIS tagged p70S6K was precipitated by Talon metal affinity resin (Clontech Laboratories, San Diego,CA). Typically, 20 III of Talon resin per 500ml of cel lysate roated at 4C for 1 hour and washed 3 times with lysis buffer. Samples were then boiled for 5 minutes in 2X gel loading dye and analyzed 34

PAGE 47

on 7.5% SOS-PAGE. Alternatively, the Talon resin/p70s6K protein complex was used in an invitro kinase assay to determine activity of p70S6K on S6 peptide. The complex was washed in kinase buffer(10mM Hepes buffer, 1 mM MgCI2, BSA, IP-20 1 mM OTT) To each sample was added 1X kinase buffer containing 1 nonradioactive ATP, 100 S6 peptide, and 200 'Y32p ATP. Samples were incubated for 15 minutes at 30 DC and the reaction stopped with 20 of 0.5M EOTA. Then 25 of the reaction mixture was loaded in duplicate on p81 phosphocellulose paper and washed 5 times with 1/200 dilution of phosphoric acid. \, The papers were dryed under IR light and counted in a liquid scintillation counter. COK1/Cyciin B Phosphorylation of Recombinant p70 86 Kinase Site specific phosphorylation of p70S6K by COK1/cyciin B complex was investigated by invitro kinase assay, and phosphorylated proteins subjected to tyrptic peptide mapping. The recombinant p70 proteins made by baculovirus expression system were employed as whole protein substrates to show differential phosphorylation by COK1/cyciin B as compared to CDK2/cyciin E. Initially, 1 X 10 6 insect cells were infected with 35

PAGE 48

either wild type, T229A mutant, or wild type deletion mutant for 48 hours, harvested, lysed in p70S6K lysis buffer, and the supernatants transferred to a new tube containing 0.1 ml of Talon purification resin. Samples were rotated at 4C for 1-2 hours. Protein complexes were centrifuged at 14,000 X g for 1 minute and washed with 1 ml of lysis buffer 3 times. Then the complexes were resuspended in 50 J.l1 of CDK kinase buffer containing 0.1 mM ATP, 200 J.lCi of ')'32p ATP, and 1 J.l1 of CDK1/cyclin B complex (MBP) or 1 J.l1 of CDK2/cyclin E. The reaction proceeded for 30 minutes and was stopped by the addition of 20 J.l1 of 0.5 M EDTA. Then 50-60ml of loading dye was added and the samples were boiled for 5 minutes. After brief centrifugation, the samples were loaded onto a 10% SDS-PAGE and electrophoresed. The gel was coomasie stained, dried, and autoradiographed. Phosphopeptide mapping Protein Precipitation and Digestion. Proteins were cut out of SDS-PAGE gels and suspended in 50 mM ammonium bicarbonate containing 1 % SOS and b-mercaptoethanol and homogenized with microcentrifuge dounce (Kontes). Samples were boiled for 5 minutes and rotated at room temperature overnight. Proteins eluted out of the gel were precipitated with 10 mg of RNAse carrier protein and 95% trichloroacetic acid for 1-2 hours. 36

PAGE 49

Proteins were pelleted by centrifugation at 1400 X g for 10 minutes, supernant drained and pellet washed with 0.4 ml of equal volume ethanol and ether and then dried for 5-10 minutes in a speedvac. Samples were treated with 20 /-11 of formic acid/ H202 (10:1) for 1-2 hours at 4 C and then 0.4 ml of water was added and the samples dried in a speedvac. Proteins were subjected to Chymotrypsin and Trypsin digestion at 1 mg/ml in NH4HC03 at 37C for 24-48 hours each. Following digestion, O.4ml of water was added and samples were heated at 65C for 10 minutes, centrifuged for 10 minutes, supernant transfered to a new tube and dried completely. Samples were washed again with 0.4 ml of water two more times and dried. Samples were subjected to Thin layer electrophoresis and Thin layer chromatography. Electrophoresis/Chromatography. Phospho-peptide samples were dissolved in 10/-11 of electrophoresis buffer ( 2.25% Formic acid, 7.75% acetic acid) and sepaprated by charge in one dimension using the Hunter thin layer peptide mapping system (HTLE-7000, CBS Scientific). For each sample, 5-10 /-11 was loaded onto TLC-cellulose plates (Merk5716 celluose, 20X20 cm, EM Science) using a blow dryer to concentrate the sample. The plate was pre-wetted with electrophoresis buffer using a soaked 37

PAGE 50

whatman paper. The samples were electrophoresed for 30 minutes at 1200 volts and plates dried for 1 hour before thin layer chromatography. The plates were placed in a TLC chamber, in an ascending position, containing chromatography buffer (65% isobutyric acid, 5% pyridine, 3% acetic acid, and 2% butanol) and allowed to run for 16-20 hours or until solvent front reached the top of the plate. Plates were then dried for 3 hours and autoradiographed. Synthetic p70 S6 Kinase Ser 411 Phospho-Peptide Map. The p70S6K peptide EPKIRSPRRFIG (Residues 406-417 of human p70S6K ) including the serine 411 phosphorylation site, was synthesized to confirm migration of the Ser411 peptide during mapping. The peptide was labelled invitro by CDK1/cyclin B kinase reaction (incorporation of y32p ATP) and the resulting phosphopeptide purified by C-18 Sep-Pak cartridge ( Millipore, Milford, MA), which removes unincorporated y32p ATP. Briefly, reaction mixture was loaded onto a prewashed column and then washed extensively with PBS until no 32p activity could be detected in the wash. The phosphopeptide was eluted out with 70% acetonitrile in a total volume of 0.1 ml and dried in a speedvac. Then 40 III of trypsin was added and digestion proceded for 24 hours. The sample was washed 3 times in water, drying in a speedvac each time. After 38

PAGE 51

final wash, the peptide was resuspended in 0.1 ml of TlE buffer and subjected to the TlE/TlC procedure. CDKlCyclin In Vitro Kinase Assay Specific activities of cdks/cyclins complexes were determined by (32p) incorporation into pH1 or p70CT peptide in the immune complex. Substrate peptides used here were; pH1, AVAAKKSPKKAKKPA (Residues 139-153 of trout Histone H1), pRB, RPPTlSPIPHIPR, and p70 CT, EPKIRSPRRFIG (Residues 406-417 of human p70S6K). The peptides were prepared by automated solid phase peptide synthesis using a type 9050 automated synthesizer from MilliGen/Biosearch, and the purity and concentration of the peptides were confirmed by HPlC. A purified Xenopus CDK1/cyclin B or a CDK2/cyciin E complex was prepared as described previously. Cells (5 x 10 6 ) were washed with PBS and lysed at 4 C in 1 ml of lysis buffer (50mM Tris-HCl, pH7.4, 1 mM EDTA, 25mM NaCI, 40mg/ml PMSF and 0.1% np40). The extract (500JlI ) was incubated for 1 hour at 4 C with either anti-CDK1, CDK2, cyclin A, or cyclin B antibody. The immune complex was absorbed to Protein Gcoupled sepharose beads for 1 hour. Alternatively, the extract was incubated for 1 hour at 4 C with p13 suc agarose beads. The beads were washed three times with lysis buffer, and once with kinase buffer (50mM Tris-39

PAGE 52

HeI, pH7.4, 10mM MgCI2, 1 mM dithiothreitol). Following the final wash, the immune complexes were suspended in 50 III of the kinase buffer containing 100mM unlabeled ATP, 200 mCi/ml [y32pJ ATP, and 10 4 M peptide. The reaction was allowed to proceed for 15 minutes at 30C and terminated by the addition of 10111 of 500 mM EDTA. Following a brief centrifugation, the supernatant (20 Ill) was applied to P-81 phosphocellulose paper (Whatman 3698-025) and radioactivity was determined using a liquid scintillation counter (Bio-Safe II, Research Products Inc., Mount Prospect, Illinois). In some experiments, a purified Xenopus CDK1/cyelin B complex or a purified Xenopus CDK2/cyclin E complex was used instead of immune complexes. Kinetic constants were determined using 108 to 103 M peptide. T cell Time Course T Cell CT-p70 S6 Kinase In Vivo Phosphorylation. T Cells were stimulated with PDB / lonomycin and harvested at 3,24,39,and 48 hours post-activation with and without rapamycin. Each time point was labelled in phosphate free dMEM containing 1 mCi of 32p inorganic phosphorous (NEN-Dupont, 1 mCi/ml, Nex-053, Boston, MA) for 3 hours prior to harvest. Cell pellets were lysed in Ripa buffer for 15 minutes. p70 S6K was immunoprecipitated by anti CT p70 antibody for 24 hours and proteins analyzed by 7.5% 40

PAGE 53

sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SOS PAGE), Coomassie stained, and autoradiographed. Activity Time Course of p70 S6 Kinase and CDK1. T cells harvested at the same time points were lysed in p70S6K lysis buffer( 10mM KP04, 1 mM EDT A, 5mM EGT A, 10mM MgCI2, 50mM b Glycerol phosphate, 1 mM Na3V04, 2mM dTT, 40mg/ml PMSF, 0.1 % NP-40) or CDK1 lysis buffer ( 50mM Tris pH7.4, 5mM EDTA, S6K 250mM NaCl, 0.1 % NP-40). COK1 and p70 was immunoprecipitated for 1 hour by anti-CDK1 antibody and anti-CT p70S6K antibody respectively,and complexed to protein G sepharose beads for 1 hour by rotating at 4C. For p70S6K activity, the beads were then washed twice in lysis buffer and once in kinase buffer (20mM MgCI2, 40mM Tris pH7.5, 2mg/ml IP-20, 0.2mg/ml BSA, 0.8mM dTT). Then 50!!1 of kinase buffer was added containing 0.1 mM ATP, 10mM S6 peptide, and 10!!Ci (10IlCi/ml) of 132 P ATP (NEN-Dupant, 30Ci/mMol) and reaction set at 30C for 15 minutes. After brief spin, 26 !!I of the reaction supernant was loaded in duplicate on p81 phospho-cellulose paper disks (Whatman 3698-025), washed in 0.5% phosphoric acid five times, and dried under infrared lamp for 15 minutes. The disks were counted in a liquid scintillation counter. For CDK1 activity, the beads were washed twice in lysis buffer and once in COK1 kinase buffer( 100 mM Tris pH7.4, 20mM MgCI2, 2mM dTT). Kinase 41

PAGE 54

reaction procedure was completed similiarly using H1 peptide as a substrate. Cell Cycle FACS Analysis. T cells collected at the same time points were analyzed to determine the percentage of cells in each phase of the cell cycle. Cells were harvested, washed in PBS, resuspended in 0.2ml of saline, and transfered to a 12X7Smm tube containing 3ml of 70% ethanol. Samples were stored at -20C overnight, centrifuged and washed with PBS. The fixed cells were incubated with O.Sml of 0.2Smg/ml ribonuclease (Sigma) in PBS at 37C for 10 minutes. The cell suspension was mixed with O.S ml of propidium iodide (Calbiochem, La Jolla, CA) solution (SOmg/ml in PBS), and after 60 minutes analyzed by a flow cytometry (EPICS Profile, Coulter) collecting red fluorescence (>600 nm) with 488 nm excitation. Increase of CT-p70 S6 Kinase Phosphorylation by Nocordizole Treatment In Vivo Cos cells at 30% confluency were set up in dMEM media and transfected with 10 mg of pCMVS-WTp70S6K plasmid using 30 ml of cell fectin reagent (Gibco-BRL) for each plate. Cells were incubated for 40 hours and then pretreated with either Nocordizole (Sigma) at SOO mm, Aphidicolin (Sigma) at 1 mm, or Rapamycin at 1 mm for S hours. Then cells were incubated for 3 42

PAGE 55

hours in the presence of 32p orthophosphate using dMEM without Na phosphate supplemented with 10% dialzed FCS. Cells were harvested and lysed in p70S6K lysis buffer for 15 minutes, centrifuged and the supernatant transfered to a new tube. AntiCT_p70S6K antibody (0.5 mg per IP) was added to the Iysates and incubated on ice for 30 minutes followed by 30 minute incubation with 30 III sepharose beads. Immune complexes were washed 3 times with p70 lysis buffer and 40% of the beads were boiled in loading dye and run on a 7.5% SOS-PAGE, the gel fixed and dried and autoradiographed. 43

PAGE 56

CHAPTER 3 Results The studies presented here investigate the possible connection between the cell cycle regulatory protein CDK1 and the major physiological kinase responsible for phosphorylating the 86 protein, referred to as p70 86 Kinase. Located in the carboxy-terminal region of the kinase are four serine/threonine phosphorylation sites, each of which are followed by a proline residue. This represents a commonly used substrate recognition motif for cyclin-dependent and mitogen-activated protein kinases. An earlier study showed that CDK1 could indeed phosphorylate C-terminal p70S6K in vitro. To further analyze this phenomenon, it was determined if other cyclin-dependent kinases could also use p70 86 kinase as a substrate or was this specific to CDK1 alone. Among the CDKs identified to date, CDK2 is the closest kinase to CDK1 by homology and substrate specificity. Therefore I initially compared CDK2 and CDK1 in the ability to phosphorylate recombinant p70S6K I developed a system to study specific phosphorylation events on the carboxy-terminus of p70S6K 44

PAGE 57

to truly analyze the activity of cell cycle regulatory proteins on this enzyme. Baculovirus Expression System Provides High Quantities of Functionally Active Wild-Type or Inactive Mutant p70 S6 Kinase for Phosphorylation Analysis In order to study phosphorylations at specific sites on p70S6K, there was a need for large quantities of native proteins that could be readily phosphorylated by various enzymes and digested with proteinases for tyrptic phosphopeptide analysis. The Baculovirus expression system provided an excellent way to accomplish these things because of the availability of highly efficient vectors for production of high quantities of heterologous proteins. The recombinant proteins contained a region upstream of the coding region consisting of six histidines which allowed for easy purification by conjugation to metal resin agarose beads. Phosphorylation of p70S6K was carried out directly on the resin beads. Two mutants were constructed which enabled analysis of the two sets of phosphorylations sites on p70 S6K. T229 A is a mutation in group 1 and a carboxy-terminal deletion mutation was also used to verify the four CT-45

PAGE 58

phosphorylation sites of group 2. Figure 5 shows the isolation and characterization of the positive clones using 8aculovirus expression vector Fastbac Htb and the restriction analysis confirming positive oreintation and correct reading frame of the transposing DNA (Figure 5a). Following transposition, Bacmid DNA was isolated which contained the cloned p70S6K of interest (Figure 5b). The subsequent expression levels of the three variants of p70 were all very high (Figure 5c) and migrated at the correct sizes. These proteins were analyzed for kinase activity and only the wild type had good activity, with the deletion mutant having partial activity and the T229A mutant nearly inactive (Figure 5d). As will be seen later, these recombinant proteins provided excellent substrates to determine differences in CDK1 and CDK2 phosphorylation of CT_p70s6K 46

PAGE 59

A) .. 8) QI .:0: .. III < z c 'C rec T229A-p70 rec l\CTp70 s:: kb 23.1 __ Bacmld DNA 9.4 1.5 6 5 4.3 2.3 2.0 C) Q D) .... Q !2 c. .... oCt c. C. 01 t!. .,.:.. C'oI :;0000 ::: C'oI () t- S5Joo t-30000 200-0 0 25)00 :x: 0 15JOO b: 10000 695000 8KG WT T229A WTDEL Figure 5. Characterization of the recombinant p70S6K clones generated by the Bac to Bac Baculovirus expression system. (A) Restriction analysis of the cDNA clones after fast Bac HTB vector ligation with p70S6K cDNA showing positive orientation. (B) Isolation of recombinant Bacmid DNA from T229A and D WT clones after transposition event with S6K vector constructs. (C) Western Blot showing expression of P70 proteins generated by Baculovirus system. (D) The kinase activities of the three enzymes, as determined by phosphorylation of the S6 peptide. See text for details 47

PAGE 60

CDK1/Cyciin B and Not CDK2/Cyciin E Phosphorylates Recombinant CT -p70 S6 Kinase on Serine 411 Phosphopeptide mapping analysis and baculovirus generated p70S6K recombinant proteins provided the tools necessary to demonstrate the specificity of CDK1/cyciin B phosphorylation on serine 411 of CT_p70S6K The rec WT-p70, rec T229A-p70, and the rec DCTp70 proteins generated by viral infection of insect cells, were complexed with Talon metal resin beads and then used as substrates for either CDK1/cyciin B (Xenopus maturation promoting factor) or CDK2/cyciin E in vitro kinase reactions. Histone H1 was also a substrate as a control of phosphorylation. The recWT -p70 protein was phosphorylated at the same intensity with and without CDK1/cyciin B (MPF) due to a high level of autophosphorylation (Figure 6a). Therefore the recT229A-p70 S6K was used as a substrate for all further phosphorylations because the group 1 T229A mutation significantly reduces the autophosphorylation of the enzyme. 48

PAGE 61

A B en .!:i "0 (.) e ...... .,.. ..... c: 8 c (.) Histone H1 recT229A-p70 w III w m .5: .5 .5 .5: (3 (3 "0 g. >>>-Q Q e Q e T"" 1: c: 0 0 0 c 0 0 (.) (.) 0 0 (.) (.) recWT-p70 recaCT-p70 w CD .5: .5 (3 g. >-"0 0 0 C\l :ro... c: 0 0 0 (.) (.) 0 recT229A-p 70 recACT-p70 Histone H1 Figure 6 One-dimensional SDS-PAGE showing in vitro phosphorylation of recombinant p70S6K proteins by CDK 1/cyclin Band CDK2Icyclin E (A) The phosphorylation of wild-type p70S6K with and without CDK1/cyclin B. (B) The phosphorylation of p70S6K mutants with CDK1/cyclin B or CDK2Icyclin E Histone H1 as a control of phosphorylation ability See text for details 49

PAGE 62

In one dimensional 7.5% 8DS-PAGE, CDK1/cyclin B phosphorylates recT229A p70S6K more intensively than CDK2/cyclin E which shows only background autophosphorylation (Figure 6b). Both CDK complexes exhibit good kinase activity as seen by the intense phosphorylation of Histone H1. As a comparison, rec DCTp70s6K showed only very slight phosphorylation without the carboxy-terminal group 2 phosphorylation sites. To further examine which serine or threonine sites on p70S6K are phosphorylated, the three bands from the recT229A 7.5% gel were cut out and subjected to trypsin and chymotrypsin digestion and the resulting phosphopeptides analyzed by the two-dimensional mapping procedure. Figure 7 shows the phosphopeptide mapping results of these three labeled bands. Neither p70S6K itself nor CDK2/cyclin E phosphorylated p70 show any specific phosphorylations with the apparent absence of carboxy-terminal phosphopeptides (Figure 7, panel A and B). CDK1/cyclin B however, exhibits good phosphorylation of the carboxy-terminal sites with one spot showing the very intense labeling (Figure 7, panel C). This spot was assumed to be the Ser 411 phosphorylation site based on the presence of the CDK1 consensus motif at that position. The serine 411 phosphorylation site (EPKIR8PRRFIG) was confirmed by the migrational analysis of synthetically made peptide representing this serine 411 peptide sequence. This phosphopeptide migrated 50

PAGE 63

to the identical position on the cellulose plate (Figure7, panel D). Taken together, these data demonstrate that among the Cterminal phosphorylation sites, the serine 411 on p70S6 kinase is most highly phosphorylated by CDK1/cycJin B in vitro. 51

PAGE 65

Figure 7. Two-dimensional thin-layer electrophoresis of the in vitro phosphorylated bands of T229A-p70s6K cut out from the one-dimensional gel seen in figure 6. Phosphorylated T229A-p70s6K proteins were subjected to chymotrypsin and trypsin digestion and resulting phospho-peptides separated by charge during thin-layer electrophoresis and then by size during thin-layer chromatography. (A) The T229A-p70s6K protein without CDK1/cyclin 8 added. (8) The phospho peptides from CDK 2/cyclin E phosphorylated T229A-p70s6K (C) Separation of phospho-peptides from CDK1/cyclin 8 phosphorylated T229A-p70s6K (D) The confirmation of the Ser 411 carboxy-terminal phosphorylation site by the similar migration of synthetically prepared Ser 411 peptide after in vitro phosphorylation by CDK1/cyclin 8. See text. 53

PAGE 66

To confirm that these phosphopeptides really represent the C-terminal phosphorylation sites, cos cells were transfected with WT_p70S6K or DCT_p70S6K (cloned into mammalian transfection vector pCMV5), labelled with 32p inorganic phosphorous and subjected to the same tryptic peptide mapping methodology. Figure 8 shows the phosphopeptide maps of p70 S6K and the individual phosphorylation sites. Panel A of figure 8 shows the migration of each phosphopeptide and the directions of TLE and TLC. The spots designated as C correspond to the carboxy-terminal sites (Ser411, Ser418, Thr421, Ser424) which were completely absent in the carboxy-terminal deletion mutant (Figure 8, panel B). Thus this methodology provides a suitable way to distinguish specific phosphorylations on the p70S6K molecule. 54

PAGE 67

recp70S6K A=T229 B=T389 C= CT peptides

PAGE 68

Figure 8. Characterization of the chymotrypsin and trypsin generated phospho peptides of p70S6K The sites marked A and B were determined to be T229 and T389 of set 1 phosphorylation sites. The C-terminal phosphorylation sites of set two are indicated by the drawn circle in the upper right quadrant of the map. The confirmation of the C-terminal phosphorylation sites is shown by the use of the C terminal deletion mutant of WT-p70S6K which shows the complete absence of the three carboxy-terminal phosphorylation spots 56

PAGE 69

The Rate of Peptide Phosphorylation by CDK1/Cyclin B Complex is Higher Than CDK2/Cyclin E Complex In agreement with my results, the region surrounding ser 411 has the best match among the C-terminal phosphorylation sites in p70S6K with the consensus substrate sequence of CDK1 (K/R-S-P-R/P-R/KlH) determined by a peptide selection approach. To further analyze the differential phosphorylation activity of CDK1 and CDK2, CTp70s6K peptide(EPKIRSPRRFIG) or histone H1 peptide(AVAAKKSPKKAKKPA) was made synthetically by amino acid synthesizer and assayed for relative rate of phosphorylation by either purified Xenopus CDK1/cyciin B complex or CDK2/cyciin E complex. Kinase reactions containing CDK complexes, 'Y 32p A TP, and 108 M to 103 M concentration of peptides were incubated for 15 minutes at 30 C. Figure 9 shows the differential enzyme kinetics of the two cyclin-dependent kinase complexes. CDK1/cyclin B complex has over ten fold higher rate of phosphorylation on CT_p70s6K peptide, with an apparent Km of 50J,.LM, than does CDK2/cyclin E complex (Km 500J,.LM). This was in contrast to the rates of phosphorylation on pH 1 (Km: for CDK1 -10 mm, for CDK2 -20mm). The specific activities were divided by 57

PAGE 70

the activity to phosphorylate 104 M of pH1 peptide to normalize. To further demonstrate the specificity of this CDK1 phosphorylation event, I also analyzed the early cell cycle proteins CDK4 and CDK6 for their potential ability to phosphorylate the C-terminus of p70S6K Figure 10a shows the rate of phosphorylation of CDK 4 and CDK 6 on p70S6K as compared to CDK1. CDK 4 and CDK 6 both show very little phosphorylation of p70S6K, however, figure 10b shows that the Immunoprecipated CDK 4 and CDK 6 proteins are active as seen by their ability to \ phosphorylate their known substrate, the whole retinoblastoma protein (Rb). 58

PAGE 71

100.________________________ 80 -8 10 cdk2/cyclin E cdc2/cyclin 8 -7 10 -6 10 -5 10 concentrations of p70CT (M) -4 10 -3 10 Fig. 9. Relative phosphorylation rate of p70S6K by purified CDK1/cyclin B or CDK2/cyclin E. A 10-a to 10 -3 molar range of C-terminal p70S6K peptide was used to assay activity of purified Xenopus CDK1/cyclin B (apparent Km of 50 !!m) or CDK2/cyclin E (apparent Km of 500 !!m)_ 59

PAGE 72

Phosphorylation rate of p70CT Phosphorylation of pRb by CDK1, CDK4 or CDK6 A 30000 B 8000 ----0---COK1 6000 COK4 20000 Ii COK6 E 4000 C-o E 2000 C-o 10000 0 e T""" v -Cl 0 c:: 0 0 () 0 0 -5 -4 -3 10 10 10 substrate concentration (M) Figure 10. Relative rate of phosphorylation on p70S6K C-terminal peptide by immunoprecipitated CDK4 and CDK6 from cycling Jurkat cells (A) A 10-5 to 10-3 molar range of substate was used, and CDK1 was used as a control. The phosphorylation activity on known substate pRB is also shown to confirm that CDK1, CDK4 and CDK6 are active complexes. 60
PAGE 73

CDK1/Cyciin B Preferentially Phosphorylates CT-p70 S6 Kinase Whereas Other Members of the Cyclin Dependent Kinase Family Do Not Since there was a significant difference in rate of phosphorylation using purified CDK complexes, the next question was to ascertain phosphorylation rates of CT_p70S6K by other members of the cyclin-dependent kinase family directly immunoprecipitated from cycling T lymphocytes. Lysates were made from T cells stimulated with PDB/ionomycin for 48 hours and treated with antibodies against CDK1, CDK2, CDK4, CDK6, cyclin A, and cyclin B. Lysates were also treated with p13suc agarose beads which immunoprecipitate both CDK1 and CDK2 and their associated cyclins. These immune complexes were conjugated to sepharose beads and assayed for phosphorylation activity on CT_p70S6K peptide EPKIRSPRRFIG by in vitro kinase assay. Phosphorylation activity of pH1 peptide and pRb peptide by these immune complexes served as a control. Table 1 shows the relative phosphorylation rates of the synthetic peptides using the various antibodies CDK1 and cyclin B more effectively phosphorylated CTp70S6K peptide than the other members of the CDK family. p13suc beads also demonstrated significant activity 61

PAGE 74

due to the presence of CDK1 in the complex. The data is presented as a percent activity to the specific activity to phosphorylate pH1 forCDK1 and CDK2; whereas the percent activity to phosphorylate pRb was used to normalize for CDK4 and CDK6. It should be remembered that CDK1 can associate with either cyclin A or cyclin 8 and that CDK2 can associate with either cyclin A or cyclin E, but only CDK1/cyciin 8 had significant activity in theses Iysates. It is unlikely that CDK1/cyciin A complex phosphorylates CT _p70S6K, but it is not clear because the ratio of CDK1 and CDK 2 in cyclin A complexes was not determined here. 80th CDK4 and CDK6 associate with the D type cyclins. Thus these data add supporting evidence that there is specific differential phosphorylation activity on the carboxy-terminus of S6K p70 by cell cycle regulatory proteins in activated T lymphocytes. Table 1. Relative phosphorylation rate of synthetic peptides by CDK1-, CDK2-, cyclin A, cyclin B, or p13 suc -immune complexes. antibodies relati ve phosphorylation rate (%) pHI p70CT anti-CDK1 100 56.3 14 2 anti-CDK2 100 5.4 4.0 anti-cyclin A 100 4.4 0.3 anti-cyc1in B 100 42.5 12 0 p13suc 100 45.9 10. 9 62

PAGE 75

Differential CDK1 Activity Phosphorylates p70 86 Kinase C Terminal Peptide at a Different Cell Cycle Phase Than Histone H1 Peptide Next we investigated where in the cell cycle the carboxy terminus of p70S6K is phosphorylated by CDK1/cyciin B. Based on the results of Table 1, p13suc beads had significant kinase activity to phosphorylate CT_p70s6K peptide due to sufficient CDK1/cyciin B activity. Therefore p13suc beads were employed to explore the phosphorylational status of CDK1 on the carboxyterminus of p70S6K peptide as compared to Histone H1 peptide during cell cycle progression. The p13suc is a CDK-associated protein which binds to CDK1 and CDK2 and possibly CDK3 and precipitate them in an active state. Lysates from human primary T cells stimulated with PDB/ionomycin were collected at selected time points of the cell division cycle and incubated with p13suc agarose beads for 1 hour and subsequently assayed for kinase activity on either histone H1 peptide or the CT_p70s6K peptide. As seen in figure 11, the activity to phosphorylate the KK8PKK peptide derived from Histone H1 was increased at about the time when CDK2 was activated in the cell cycle corresponding to the G1/S phase. In contrast, the activity to phosphorylate the KIRSPRR peptide derived from CT-p70S6 kinase was increased later in the cell cycle, rather in parallel to CDK1 activation at the G2/M boundary. These results suggested that the cyclin-63

PAGE 76

dependent kinase activity on CT_p70S6K was distinct from the histone H1 phosphorylation earlier in the cell cycle. 8000 ____________ --, 6000 4000 2000 Substrate Peptide --0-pH1 p70CT o o 6 12 18 24 30 36 42 48 hr T cells stimulated with PDBullonomycin GO/G1 /8 /{21M I Fig. 11 p13suc-associated kinase activItIes to phosphorylate pHI 0 r C-terminal P70S6K peptide. Stimulated T lymphocytes were harvested at sequential times throughout the cell cycle. CDKlIcyclin B was immunoprecipitated using p13suc beads and the kinase activities of each time point assayed with either pHI or p70CT as a substrate. FACS analysis was done to confirm cell cycle timing. 64

PAGE 77

In summary, the in vitro studies presented evidence of differential cyclin-dependent kinase activity on a specific peptide sequence located in the carboxy-terminus of p70S6K that demonstrated over 10 fold higher affinity for CDK1/cyciin B than CDK2/cyciin E. This was shown to be a phenomenon more exclusively associated with CDK1 than other members of the cye/in-dependent kinase family such as CDK2, 3, 4, and 6; and this phosphorylation activity correlated with the G2\M phase of the cell cycle, distinct from the more common substrate histone H1 which becomes phosphorylated around the G1/8 transition. These in vitro findings established the rationale to search for evidence that this was occurring in vivo and had physiological significance. The focus shifted to exploring phosphorylation events occurring in actively cycling cells in hopes of finding further evidence linking components of the cell cycle regulatory system to a key enzyme of an important cell signal transduction pathway. Late Cell Cycle Phosphorylation Versus Early Activity of p70 86 Kinase CDK1 is known to be highly abundant and active late at the cell cye/e in the G2 and M phases. In vitro evidence presented in 65

PAGE 78

Figure 12 suggested CDK phosphorylation on CT_p70S6K peptide was occurring late in the cell cycle at the G2/M boundary Therefore, I explored a potential link in the late cell-cycle phosphorylation of the carboxy-terminal region of p70S6K and components of the cell cycle machinery at the G2/M phase in vivo. While investigating in vivo time course activity of p70S6K it became clear that their was a discrepancy in the early and rapid activation of p70S6K in the group 1 phosphorylation sites and the late cell cycle phosphorylation of carboxy-terminal group 2 phosphorylation sites. The activities of p70S6K and CDK1 as they correlate to the cell cycle are shown in figure 12b and 12c. p70S6K activity is most active very early after primary T cells become activated by PDB/ionomycin and continually declines through progression of the cell cycle. In contrast, CDK1 activity is maximal very late in the cell division cycle corresponding to the G2/M boundary (Figure 12c). The immunosuppresant drug Rapamycin totally inhibits the p70S6K activity, however the late activity of CDK1 is rather resistant to Rapamycin. FACS analysis confirms cells were entering G21 M phase only during the latter parts of the time course (Figure 12d). In contrast to the early activity of p70S6K the enzyme became highly phosphorylated very late in cell cycle progression in vivo. Figure 12a shows T cells stimulated with PDB/ionomycin and labeled with inorganic 32p phosphorous for 3 hours prior to each harvest. 66

PAGE 79

Immunoprecipitations with anti-CT p70 antibody show that p70S6K was intensively phosphorylated late in the cell cycle and that this in vivo phosphorylation was largely resistant to Rapamycin treatment indicating a possible connection to group 2 phosphorylation sites in the carboxy-terminus. At the protein level, Figure 12a (second photo of Figure 12a) shows Western blots indicating that p70S6K was indeed activated by PDB\I as seen by the upper band at early time points, and this was inhibited by the addition of Rapamycin. Total protein amount was similar for each sample load. 67

PAGE 80

PDBuII PDBuII +RAP 0 3 24 39 48 3 24 39 48 hours (A) 200,... 97-69-p70S6K 46-p70S6K -lgHC -(8) E 50000 e. 0 40000 >--"S 30000 .. 0 as 20000 JI:: CD 10000 en 0 ..... e. 0 10000 (C) -E e. 8000 U >-6000 -oS .. 4000 0 as 2000 C 0 0 (0) 120 s::: 100 0 80 ;; co 0 GO/G1 ::s 60 c. m1 s "'" 0 40 G2IM e. eft. 20 0

PAGE 81

F ure 12. The phosphorylation of p70S6K in cycling human T lymphocytes by 32p ophosphate labeling in vivo. T cells were stimulated with PD8110nomycin +/R pamycin, radioactively labelled with 32p phosphorous and collected at selected time p? nts throughout the cell cycle. Each time point was labelled for 3 hours. The p70S6K Pf teins were immunoprecipitated with anti-G-terminal p70S6K antibody conjugated to sf. harose and separated on a 7.5% polyacrylamide gel. (A) The results showing 32p I elled p70S6K during the time course. Protein amount was nearly equal. (8) The ki, se activity of the p70S6K present in each time point, assayed on the phosphorylation 01 6 protein in vitro. (G) The kinase activity of GDK1 present in each time point, as ayed on the phosphorylation of Histone H1 protein in vitro. (D) FAGS staining wing the cell cycle timing of each sample, confirming that the late time point cells w e in the late phases of the cell cycle. 69

PAGE 82

In a similiar experiment, T lymphocytes were treated with Rapamycin only during 45-48 hours after stimulation with PDB/lonomycin instead of stimulation from the beginning of the time course. This much more specifically timed addition of Rapamycin also showed no effect on the phosphorylation of p70S6K as shown in figure 13. This reiterates the possibility that this phosphorylation of p70S6K late in the cell cycle is principally due to the phosphorylation occuring at the carboxy terminus. Protein amounts for each lane were relatively constant. 70

PAGE 83

c..
PAGE 84

The Enhancement of CDK1 Protein Expression and Activity With Nocordizole Treatment Also Markedly Increases p70 86 Kinase Phosphorylation In Vivo The T cell time course experiments revealed a distinct late cell cycle phosphorylation with potent CDK1 activity correlating to the same phase of the cell cycle But was CDK1 actually phosphorylating CT _p70S6K in vivo? To provide some initial supporting evidence I used 32p labeled transfected cos cells to study the phosphorylational status of p70S6K The treatment of these cells with known modulators of the cell cycle like Nocodizole, Aphidicolin, and Rapamycin, made it possible to inquire about specific p70 phosphorylations at selected times in the cell cycle. Cos cells transfected with p70WT-CMV5 (a mammalian transfection vector containing the sequence of wildtype p70S6K cDNA) were pretreated 5 hours with the drugs prior to harvest. Immune complexes constructed with CT-p70 antibody were made from these cells and assayed for p70 S6K phosphorylational status. Nocodizole (an inhibitor of microtubule assembly) stops the cell division cycle in the G2/M phase and dramatically elevates the level of CDK1 in the cell (Figure 14). Aphidicolin is an inhibitor of DNA polymerase alpha and delta, and stops the cell cycle at the G1/S phase and only slightly altered 72

PAGE 85

CDK1 activity. Rapamycin, a specific inhibitor of p70S6K had no effect on CDK1 activity, however, it completely inhibited p70S6K activity (Figure 14d). The activity of p70S6K was only moderately affected by Nocodizole and Aphidicolin. With the dramatic increase in CDK1 activity by Nocodizole treatment, the phosphorylation of p70S6K was also increased while actual protein amounts remain equal (Fi gure 14a and 14b). These results add supporting evidence that CDK1 may, indeed, be involve in the phosphorylation of p70S6K during the G2/M transition in vivo. 73

PAGE 86

A B C e a. u "> ;: u ca .:tt. CD (h <:) 1'0 a. 0 e a. u >"> :g ca C 0 kD con. nco aph rap 97 46 500000 400000 300000 200000 100000 0 300000 .200000 100000 0 (5 "'" .... c 0 0 0 0 0 Z -p70S6K IgHC p70s6k CDK1 l: C. c.
PAGE 87

Figure 14. Cos cells transfected with p70S6K WT and cultured in the presence of 3 2p orthophosphate and either Nocordazole, Aphidicolin, or Rapamycin for 3 hours in vivo. Cells were harvested and p70S6K immune complexes were analyzed by 7.5% gel (A). Nocordazole increases p70S6K Rapamycin resistant phosphorylation. Western blotting of the same gel to show equal protein amount (8). Each sample was analyzed for P70S6K activity in vitro (C), and CDK1 activity in vitro(D). CDK1 activity is very high in Nocordazole sample, whereas p70S6K activity does not correlate with this in vivo phosphorylation 75

PAGE 88

CHAPTER 4 DISCUSSION The investigations presented here demonstrate phosphorylation activity on C-terminal p70S6K by CDK1 and not by other members of the cell cycle regulatory machinery. This is counterintuitive because p70S6K functions early in the cell cycle during G1, and the CDKlcyclin pairs that should correlate with S6K -p70 should be CDK4 and CDK6 complexes because they function early in G1. This correlation should occur at the group 1 phosphorylaton sites. But substantial evidence is presented here showing a specific phosphorylation on serine 411 in the group 2 phosphorylation site by CDK1/cyclin B in vitro. Indeed, in a recent study of the substrate specificity of CDK1 and CDK2 KSPRK sequence (closely matching Histone H1 peptide that I used) is equally phosphorylated by CDK1/cyclin B, CDK1/cyclin A, CDK2/cyclin A and CDK2/cyclin E (Holmes, J. K., and M. J. Solomon. 1996) In contrast however, when the last K amino acid is substituted to R (making a peptide sequence very close to p70S6K peptide I used), CDK2/cyclin A and CDK2/cyclin E do not 76

PAGE 89

phosphorylate the peptide %) whereas CDK1\ cyclin B maintains about 70% of the activity to phosphorylate the original peptide. This suggests that the R414 amino acid in the sequence R8PRR of p70S6K is responsible for the preference of CDK1 as the kinase to phosphorylate p70S6K Given that CDK1 correlates with the G2/M phase of the cell cycle, it is not surprising then that the specific phosphorylation of CT_p70s6K by CDK1 occurs predominantly late in the cell cycle. Preliminary evidence presented here indicates that this phosphorylation event may be occurring in vivo, and that this may be important for transition into mitosis. The discrepancy in the early activity of p70S6K and the late cell cycle phosphorylaton of the C-terminus indicates two possible mechanisms regulating p70S6K ; one Rapamycin sensitive and responsible for the enzyme's principal activity, and one regulating another physiologically important interactions. Implications for CDK1 Phosphorylating p70 86 Kinase Taken together theses findings present an interesting question. What is CDK1/cyciin B doing functionally to the C terminus of p70S6K when it phosphorylates it and why is this event necessary late in the cell cycle instead of early in G1? One possibility could be that the kinase is rendered inactive by this massive phosphorylation in the C-terminus. CDK1 is known to 77

PAGE 90

phosphorylate and inactivate many proteins involved with cytoskeletal reorganization during mitosis. However, we and others have shown that when the C-terminus of p70S6K is mutated or deleted, the kinase still exhibits activity due to the group 1 phosphorylation site in and around the catalytic site. Alternatively, the kinase could be rendered constitiutively active by altering the interaction between the pseudosubstrate in the carboxy-terminus and the amino terminal acidic region. This would alter the ability of the kinase to be activated correctly, leading to an enzyme with the active site continually exposed, and thereby leaving p70S6 K in a hyperactive state until after mitosis is complete. However in vivo data presented here indicates p70S6K activity is seen only early after cell stimulation with very low kinase activity at the time this C-terminal phosphorylation occurs (see Figure 1 b). Therefore, some other function must be bestowed on the kinase to account for this phosphorylation event. Subcellular Localization It can be speculated that there is a novel function for this kinase late in the cell cycle elicited by the interaction of specific binding proteins in and around the C-terminus of p70S6K Specific binding proteins altered by C-terminal phosphorylation may in turn impart new physiological responses required for 78

PAGE 91

division via differential subcellular localization of the kinase. Preliminary immunofluorescent labeling studies indicate subtle changes in subcellular localization of p70S6K in the G2/M phase, with p70S6K co-localizing with actin filaments and some differential staining showing a more perinuclear localization in G2/M (unpublished observations). These binding proteins may activate other biochemical pathways in the cytoplasm and subsequently impart changes in the nucleus. Nucleolar Organization The nucleus in fact offers some other intriguing speculation. Thomas et al. have demonstrated a distinct pool of phosphorylated S6 in the nucleus. The absence of protein synthesis in the nucleus suggests a unique role for this kinase and this pool of nuclear phosphorylated S6. The nuclear isoform of this kinase, p8SS6K (containing the 23 aa nuclear localization region in N-terminus), is expressed at a level 100 fold less than p70S6K It could be speculated that the subcellular localization changes induced by C-terminal phosphorylation of p70S6K may affect or augment the phosphorylation of nuclear S6, which may be involved in regulating ribosome biogenesis in the nucleolus. During prophase of mitosis, the nucleolus first decreases in size then dissociates. In metaphase, the nucleolus disappears 79

PAGE 92

completely, and when ribosomal RNA synthesis restarts in telophase, small nucleoli reappear at the chromosomal locations of the ribosomal RNA genes (Miller, O.L. 1981; AnastassovaKristeva, M. 1977). It may be postulated that p70/85s6K plays an important function when the nucleolus is disintegrated and distributed over all metaphase chromosomes to be carried to each of the new daughter cells and regenerated into new nucleoli. Future Experiments Several experiments will be conducted in our lab to address some of these issues. Initially, I will further demonstrate the C terminal phosphorylation of p70S6K by CDK1 \ cyclin B in vivo by endoproteinase Glu-C (V8 protease) digestion of p70S6K from actively cycling cells. The V8 protease isolates the specific C terminal region that was studied in vitro, so that this phosphorylation event can be confirmed to be physiologically important in the intact cell. Also I plan to construct a CDK1 dominant negative mutant and analyze the effects this may have on the phosphorylation of C-terminal p70S6K in vivo to further clarify the importance of this interaction. Additional subcellular localization studies will be conducted with Nocodazole treated cells to examine the effect of increasing the level of CDK1 expression on p70S6K function and its location in the cell. Since 80

PAGE 93

there are two distinct activational mechanisms assumed to be involved in the regulation of this kinase, it would be interesting to explore other potential interactions by members of the MAPK like kinases. As stated earlier, the motifs in the C-terminus of p70S6K can be recognized by MAPK-like enzymes. In our lab, there is preliminary evidence that MEKK3, an isoform of MEKK1, upregulates p70S6K phosphorylation ( Unpublished observations). MEKK2 and MEKK3 are newly identified kinases similiar in structure to each other but different than MEKK1, that are proposed to be involved in the p42 JUN kinase and the p38 MAP kinase pathway, but their function and proper placement in the signaling cascade remains to be elucidated. If there is a link between MEKK3 and C-terminal phosphorylation of p70S6K this may add another level of regulation for this kinase. Whether or not this means anything in the context of regulation of growth and proliferation remains to be investigated. 81

PAGE 94

APPENDIX Equipment DNA Thermal Cycler: Perkin Elmer Cetus, Inc Norwalk, CT Water Jacketted Incubators: Forma Scientific a division of Mallinckrodt Inc. Marietta, OH Mammalian cells 37C 5% C02 Insect cells 30C 5% C02 Laminar Flow Culture Hood: Biogard Hood Baker Company Sanford, Maine Scintillation Counter: Beckman LS 5000 T A Beckman Scientific Fullerton, CA Spectrophotometer: Beckman DU-65 Beckman Scientific Fullerton, CA Cell Counter and Channelyser: Coulter Counter 2M Coulter Scientific Miami, FL Power Supply: Biorad Model 3000Xi 82

PAGE 95

BioRad, Hercules, CA Centrifuges: Large Volume; GPKR centifuge, Swinging bucket 8000 RPM Beckman, Fullerton, CA Small volume; Marthon 13K1M 12,000 RPM Fisher Scientific Pittsburgh, PA Speed Vacuum: Speed vac plus SC110A Savant, Farmingdale NY Gel dryer: Drygel Sr., Slab Gel Dryer Model 1160 Hoefer, Scientific San Fransisco, CA Film and Cassettes: New England Nuclear Life Sci. Products Reflection Autoradiography film #679062 8X10 In. Cassettes; Fisher, Biotech. FBAC 810 Autoradiograghy cassette Protein Electrophoresis and Blotting: Mini-Protein II Electrophoresis Cell 1.0 or 1.5 mm thick spacers, 10 well BioRad, Hercules, CA MiniTrans-Blot Electrophoresis Transfer Cell BioRad, Hercules,CA Nitrocellulose: 83

PAGE 96

Hybond, ECl Amersham, Life Sciences Arlington Heights, Il Two-Dimensional electrophoresis: HTlE-7000 Hunter Thin-layer Electrophoresis System C.B.S. Scientific, Del Mar, CA Cellulose Plates: DC-Fertigplates, Cellulose, EM-5716-7 VWR, Denver, CO CHEMICALS Fisher Scientific Methanol A408-4 Pyridine BP-1155-500 Hydrogen peroxide H323-500 2-Propanol A451-1 Ethyl Ether E138-1 Acetic acid A38c-212 Formoc acid A 119-1 Potassium phosphate dibasic 3252-1 Potassium phosphate monobasic 3246-01 Magnesium chloride M33-500 Sigma. ST. louis. MO Phosphoric acid P6560 2-Mecapto Ethanol M6250 laurel Sulphate l4390 Sodium Chloride S9888 Citric Acid C8532 Trizma Base T8524 Ammonium Bicarbonate A6141 84

PAGE 97

EGTA E4378 EDTAEDS Beta Glycerol phosphate G6376 Tricine T7911 Triton X-100 T8678 Trichloro acetic acid T -8657 Sodium orthovanadate S6508 Potassium chloride P4504 Isobutyric acid 11754 Alpha Chymotrypsin C3142 dTT (dithiothreital) Worthington Biochemical Corp .. Freehold. NJ TrypsinTPCK treated Boehringer Mannheim Protein gel mix 168582 Seakem agarose 50074 Promega Biochemical. Madison WS Neuramindase 480717 Gemini L-Glutamine 400-110 Pen/strep Calbiochem Deoxycholic acid 264101 NP-40 492017 85

PAGE 98

Pharmacia Ficoll-Paque 17-084003 BioRad Ammonium Persulphate 161-0700 Gibco-BRL. Gaithersburg, MD SF9-900 II SFM media 10902-096 Temed 15524-010 RPMI 1640 media 21870-076 Dulbeccos Modified Eagle media 11855-084 Sodium phosphate 11971-028 Opti-MEM 11057-021 Hepes Buffer 15630-080 Summit Biotech, Fort Collins. CO SerumFetal Bovine Serum Radioactivity NEN Dupont. Boston. MA Easy tide BLU-NEG-502X gama 32P ATP 10mci/ml Phosphorous 32 inorganic NEX-053 86

PAGE 99

CELLS Human Perpheral Mononuclear Cells Childerens Hospital Blood donor center/ Aphoresis unit Denver, CO Cos-7 cells Kidney, SV-40 transformed, African Green Monkey cells ATCC (American type Culture Collection) Cold Spring Harbor NY Insect cells SF9 cells (spodoptera frusipeda) fall armyworm larval cells Gibco-BRL Gaithersburg, MD Sheep Red Blood cells Colorado Serum CO Denver, CO REAGENTS Talon Metal Affinity Resin Clontech 8901 Nickel NTA resin YO 2281, Quiagen Cell fectin reagent, Gibco-BRL 10362-010 Xenopus CDK1/cyclin Band CDK2/cyclin E Jim Maller University of Colorado Health Sciences Center Denver, CO Histone H1 protein, Boehringer Mannheim, CA 1004875 Rainbow Marker, RPN 756 Amersham, Arlington Heights, IL 87

PAGE 100

Bac to Bac Baculovirus expression system Gibco-BRL, pfast HT expression vectors 10584-019 rec-Protein G 8epharose 4B 10-1242 Zymed, 8an Fransisco, CA SOLUTIONS Ripa buffer 50 mM Tris (pH 7.5) 150 mM NaCI 1% NP-40 0.5% sodium deoxycholate 0.1%8D8 p70S6K lysis buffer 10 mM KP04 1 mM EDTA 5 mM EGTA 10 mM MgCI2 50 mM beta Glycerol phosphate 1 mM sodium vanadate 2mM dTT 40 ug/ml PM8F 0.1% NP-40 0.5% Aprotinin 0.5% Leupeptin p70S6K kinase buffer 20 mM MgCI2 40mM Tris (pH7.5) 2 mg/ml IP-20 0.2 mg/ml B8A 0.8 mM dTT 88

PAGE 101

Electrophoresis buffer 25 mM Tris base 0.2M Glycine 3 mM SOS Transfer buffer 25 mM Tris base 0.2 M Glycine T8ST buffer 1.5 M NaCI 0.5 M Tris (pH 7.4) 0.5% Tween 20 89

PAGE 102

References Anastassova-Kristeva, M. The nucleolar cycle in man. J. Cell Sci. 25:103-110, 1977. Ballou, L. M., Luther, H. Thomas, G. 1991. MAP2 kinase and 70K S6 kinase lie on distinct signaling pathways. Nature 349:348. Bandi, H. R., S. Ferrari, 1993. Identification phosphorylation sites stimulated with serum. J. Krieg, H. E. Meyer, and G. Thomas. of 40S ribosomal protein S6 in Swiss mouse 3T3 fibroblasts J. BioI. Chern. 268:4530-4533. Banerjee, P., M. F. Ahmad, J. R. Grove, C. Kozlosky, D. J. Price and J. Avruch. 1990. Molecular structure of a major insulin/ mitogen-activated 70-kDa S6 protein kinase. Proc. Nat/. Acad. Sci. USA 87:8550. Bierer, B. E., Mattila, P. S., Standaert, R. F., Herzenberg, L. A., Burakoff, S. J., Crabtree, G., and Schreiber, S. L. 1990. Two distinct signal transmission pathways in T lymphocytes are inhibited by complexes formed between an immunophilin and either FK506 or rapamycin. Proc. Natl. Acad. Sci. USA. 87, 9231-9235. Blenis, J., Chung, J., Erickson, E., Alcorta, D. A., Erickson, R L. 1991. Distinct mechanisms for the activation of the RSK kinases/pp90rsk and pp70-S6 kinase signaling systems are indicated by inhibition of protein synthesis. Cell Growth Differ. 2:279. Bommer, U.-A., Noll, F., Lutsch, G., and Bielka, H. 1980. Immunochemical detection of proteins in the small subunit of rat liver ribosomes involved in binding of the ternary initiation complex. FEBS Lett. 111, 171-174. 90

PAGE 103

Bos, J.L. 1995. A target for phosphoinositide 3-kinase: AktlPKB. TIBS 20, 220-221 Boudewijn M, Burgering T, and Coffer, P. J. 1995 Protein kinase B (c-Akt) in phosphatidyl-3-0H kinase signal transduction. Nature 376 599-602. Brown E. J., Alberts, M. W., Shin, T. B., Ichikawa, K., Keith, C. T., Lane, W. S. and Schreiber, S. L. 1994. A mammalian protein targeted by G1-arresting rapamycin-receptor complex. Nature 369, 756-758. Brown, E. J., P. A. Beal, C. T. Keith, J. Chen, T. B. Shin and S L. Schreiber. 1995. Control of p70 S6 kinase by kinase activity of FRAP in vivo. Nature 377:441. Burgering, B. M. T. and P. J. Coffer. 1995. Protein kinase B (c-Akt) in phosphatidylinositol-3-0H kinase sig nal transduction. Nature 376:599. Calvo, V., Crews, C. M., Vik, T. A., and Bierer, B. E. 1992. Interleukin 2 stimulation of p70 86 kinase activity is inhibited by the immunosuppressant rapamycin. Proc. Natl. Acad. Sci. USA 89, 7571-7575. Cheatham, B., Vlahos, C. J., Cheatham, L., Wang, L., Blenis, J., and Kahn, C. R. 1994. Phosphatidylinositol 3-ki nase activation is required for insulin stimulation of pp70 86 kinase, DNA synthesis, and glucose transporter translocation. Mol. Cell. BioI. 14, 4902-4911. Chou M. and Blenis, J 1995. The p70kDa 86 Kinase: regulation of a kinase with multiple roles in mitogenic signalling. Curro Opin. Cel. Bioi. 7, 806-814. Chung, J., Kuo, C. J., Crabtree G. R., and Blenis, J. 1992. Rapamycin-FKBP specifically blocks growth-dependent activation of and Signaling by the 70 kd 86 protein kinases. Cell 69, 1227-1236. 91

PAGE 104

Chung, J., Grammer, T. C., Lemon, K. P., Kazlauskas, Blenis, J. 1994. PDGFand insulin-dependent activation mediated by phosphatidylionsitol-3-0H Nature 370, 71-75. A., and pp70S6K kinase. Coffer, P. J., and Woodgett, J. R., 1994. Differential subcellular localization of two isoforms of p70S6 protein kinase. Biochem. Biophys. Res. Commu 198, 780-786 Dennis, P.B., Pullen, N., Kozma, S.C. and Thomas, G. 1996. The principal Rapamycin-sensitive p70 S6 Kinase phosphorylation sites, T-229 and T-389, are differentially regulated by Rapamycin-insensitive kinase kinases. Mol Cel. BioI. 16, 6242-6251. Downward, J. 1995. A target for PI(3) Kinase. Nature 376, 553-554. Elledge, S. J. and Spottswood, M. R. 1991. A new human p34 protein kinase, CDK2, identified by complementation of a cdc28 mutation in Saccaromyces cerevisiae, is a homolog 0 f Xenopus Eg1. EMBO J. 10:2653-2659. Ferrari, S., R. B. Pearson, M. Siegmann, S. C. Kozma and G Thomas. 1993. The immunosuppressant rapamycin induces inactivation of p70S6K through dephosphorylation of a novel set of sites. J. BioI. Chem. 268:16091. Ferrari, S., W. Bannwarth, S. J. Morley, N. F. Totty and G Thomas. 1992. Activation of p70S6K is associated wit h phosphorylation of four clustered sites displaying Ser/ThrPro motifs. Proc. Nat!. Acad. Sci. USA 89:7282. Franco, R., and Rosenfelf, M. G. 1990. Hormonally inducible phosphorylation of a nuclear pool of ribosomal protein S6. J. BioI. Chem 265, 4321-4325. Geyer, P. K., Meyuhas, 0., Perry, R. P., and Johnson, L. F. 1982. Regulation of ribosomal protein mRNA content and 92

PAGE 105

translation in growth-stimulated mouse fibroblasts. Mol. Cel. BioI. 2, 685-693. Grove, J.R., Banerjee, P., Balasubramanyam, A., Coffer, P.J., Price, D.J., Avurich, J. and Woodgett, J.R. 1991. Cloning and expression of two human p70 S6 Kinase polypeptides differing only at their amino termini. Mol. Cel. BioI. 11, 5541-5550. Han, J. W., R. B. Pearson, P. 8. Dennis and G. Thomas. 1995. Rapamycin, wortmannin, and the methylxanthine SQ2006 inactivate p70S6K by inducing dephosphorylation of the same subset of sites. J. BioI. Chern. 270:21396. Hanks, S. K., Quinn, A. M., and Hunter, T. 1988. The protein kinase family: Conserved features and deduced phylogeny 0 f the catalytic domains. Science 241 :42-52. Holmes, J. K., and M. J. Solomon. 1996. A predictive scale for evaluating cyclin-dependent kinase substrates: A comparison of p34cdc2 and p33cdk2. J. BioI. Chern. 271 :25240-25246. Jefferies, H. B. J., C. Reinhard, S. C. Kozma and G. Thomas. 1994. Rapamycin selectively represses translation of the polypyrimidine tract mRNA family. Proc. Nat!. A cad. Sci. USA 91 :4441. Jeno, P., Ballou, L. M., Novak-Hofer, I., and Thomas, G. 1988. Identification and characterization of a mitogen-activated S6 kinase. Proc. Natl. Acad. Sci. USA 85:406-410 Kjeldgaard, N. O. and Gausing, K. 1974. Regulation 0 f biosynthesis of ribosomes. in Ribosomes, Nomura, M., Tissieres, A. and Lengyel, P. ed., Cold Spring Harbor Laboratory, pp369-392. Kitagawa, M., H. Higashi, H-K. Jung, I. Suzuki-Takahashi, M Ikeda, K. Tamai, J. Katao, K. Segawa, E. Yoshida, S. 93

PAGE 106

Nishimura ands Y. Taya. 1996. The consensus motif for phosphorylation by cyclin D1-cdk4 is different from that for phosphorylation by cyclin A/E-cdk2. EMBO J 15:7060-7069. Kozma S. C., Ferrari, 8. and Thomas, G. 1989. Unmasking a growth factor/oncogene-activated 86 phosphorylation cascade. Cell Signalling 3, 219-225. Kuo, C. J Chung, J., Fiorentino, D. F., Flanagan, W. M., Blenis, J., and Crabtree, G. R. 1992. Rapamycin selectively in h i bits interleukin-2 activation of p70 86 kinase. Nature 358, 7073. Lane, H. A., A. Fernandez, N. J. C. Lamb, and G. Thomas. 1993. p70S6K function is essential for G1 progression. Nature 363:170. Levy, S., Avni, D., Hariharan, N., Perry, R. P., and Meyuhas, 0. 1991. OJigopyrimidine tract at the 5' end of mammlian ribosomal protein mRNAs is required for their translational control. Proc Nat!. Acad. Sci. U8A 88, 3319-3323. Miller, O.L. The Nucleolus, chromesomes, and visualization of genetic activity. J. Cell BioI. 91 :15s-27s, 1981. Ming, X.-F., B. M T. Burgering, 8. Wennstrom, L. ClaessonWelsh, C.-H. Heldin, J. L. Bos, 8. C. Kozma and G Thomas. 1994. Activation of p70/ p85 86 kinase by a pathway independent of p21 ras. Nature 371 :426. Morris, R. E. 1991. Rapamycin: FK506 s fraternal twin 0 r distant cousin? Immunol Today 12, 137-140. Mukhopadhyay, N. K., D. J. Price, J. M. Kyriakis, 8. Pelech J 8anghera, and J. Avruch. 1992. An array of insulinactivated, proline-directed serine/threonine protein kinases phosphorylate the p70 86 kinase. J BioI. Chem. 267:3325-3335. 94

PAGE 107

Nakanishi, T., Kohno, K., Ishiura, M., Ohashi, H., and Uchida, T. 1988. Complete nucleotide sequence and characterization of the 5 -flanking region of mammalian elongation factor 2 gene. J. BioI. Chem. 263, 6384-6391. Nigg, E. A. 1993. Targets of cyclin-dependent kinases. Curro Opin. Cell BioI., 5:187-193. Patel, H. R., and Miller, R. A. 1991. Analysis of protein phosphorylation patterns reveals unanticipated com plexity in T lymphocyte activation pathways. J. Immunol. 146, 3332-3339. Pearson, R. B., P. B. Dennis, J.-W. Han, N. A. Williamson, 8. C. Kozma, R. E. H. Wettenhall, and G. Thomas. 1995. The principle target of rapamycin-induced p70S6K inactivation is a novel phosphorylation site within a conserved hydrophobic domain. EMBO 14:5279-5287. Pines, J. 1993. Cyclins and cyclin-dependent kinases: take your partners. TIBS 18, 195-197. Price, D. J., Grove, J. R., Calvo, V., Avruch, J., and Bierer, B. E. 1992. Rapamycin-induced inhibition of the 70-kilodalton 86 protein kinase. Science 257, 973-977. Proud, C. G. 1996. p70 86 kinase: an enigma wit h variations. TlBS 21 :181-185. Reinhard, C., Fernandez, A., Lamb, N. J. C., and Thomas, G 1994. Nuclear localization of p85s6K: functional requirement for entry into 8 phase. EMBO J. 13, 1557-1565. Reinhard, C., G. Thomas, and 8. C. Kozma. 1992. A single gene encodes two isoforms of the p70 86 kinase: Activation upon mitogenic activation. ProC. Nat!. Acad. Sci. USA 89:4052-4056. 95

PAGE 108

Schreiber, S. L. 1991. Chemistry and biology of the immunophilins and their immunosuppressive ligands. Science 251 :283-287. Sherr, C.J. Mammalian G1 Cyclins. 1993. Cell. 73:1059-1065. Sugiyama, H., P. Papst, E. W. Gelfand, and N. Terada. 1996. p70 S6 kinase sensitivity to rapamycin is eliminated by amino acid substitution of Thr229 J. Immunol. in press. Susa, M., A R. Olivier, D. Fabbro, and G. Thomas. 1989. EGF induces biphasic S6 kinase activation: late phase is protein kinase C-dependent and contributes to mitogenicity. Cell 57:817-824. Terada, N., H. R. Patel, K. Takase, K. Kohno, A C. Nairn and E W. Gelfand. 1994. Rapamycin selectively inhibits translation of mRNAs encoding elongation factors and ribosomal proteins. Proc. Natl. Acad. Sci. USA 91 :11477. Terada, N., Lucas, J. J., Szepesi, A., Franklin, R A, Domenico, J. and Gelfand, E. W. 1993. Rapamycin blocks cell cycle progression of activated T cells prior to events characteristic of the middle to late G1 phase of the cycle. J. Cell. Physiol. 154, 7-15. Terada, N., J. J. Lucas, A. Szepesi, R. A. Franklin, K. Takase and E. W. Gelfand. 1992. Rapamycin inhibits the phosphorylation of p70 S6 kinase in IL-2 and mitogen activated human T cells. Biochem. Biophys. Res. Commun. 186:1315. Terada, N., K. Takase, P. Papst, A. C. Nairn and E. W. Gelfand. 1995. Rapamycin inhibits ribosomal protein synthesis and induces G1 prolongation in mitogen-activated T lymphocytes. J. Immunol. 155:3418. 96

PAGE 109

Terada, N., R. A. Franklin, J. J. Lucas, J. Blenis and E. W. Gelfand. 1993. Failure of rapamycin to block proliferation once resting cells have entered the cell cycle despite inactivation of p70 S6 kinase. J. BioI. Chern. 268:12062. Weng, Q.-P., K. Andrabi, A. Klippel, M. T. Kozlowski, L. T. Williams and J. Avruch. 1995. Phosphatidylinositol-3-kinase signals activation of p70 S6 kinase in situ through site-specific p70 phosphorylation. Proc. Nat!. Acad. Sci. USA 92:5744. Weng, Q.-P., K. Andrabi, M. T. Kozlowski, J. R. Grove and J. Avruch. 1995. Multiple independent inputs are required fo r activation of the p70 S6 kinase. Mol. Cell. BioI 15:2333. Zhang, J., R. J. Sancez, S. Wang, C. Guarnaccia, A. Tossi, S. Zahariev, and S. Pongor. 1994. Substrate specificity of cdc2 kinase from human Hela cells as determined with synthetic peptides and molecular modeling. Archives of Biochemistry and Biophysics 315(2):415. 97