Maturation Promoting Factor MPF

1
(No Transcript)
2
Maturation Promoting Factor (MPF)
3
Properties of MPF
4
Genetics identified the Cdc2/Cdc28 protein
kinase, functionally conserved from yeast to
humans
S. pombe fission yeast
S. cerevisiae budding yeast
cdc2ts at 37oC
cdc28ts at 37oC
5
Cyclin
6
MPF cyclin-Cdc2 CDK activity low in G1, high in
S/G2/M
7
Multiple layers of regulation control CDK activity
M/G1
G1/S
mitotic entry
8
Structural basis of CDK regulation
cyclin
CDK
9
CDK Inhibitors (CKIs) are critical for CDK
regulation
p27KIP1
10
The budding yeast cell division cycle
pheromone
nutrients
ploidy
START commitment point (restriction point)
G1
M
S

• bud emergence
• SPB duplication
• DNA replication
• cyclin dependent
• kinases ON

G2
L. Hartwell
11
Cyclin-dependent kinases (CDKs) are activated by
cyclins and inhibited by CDK inhibitors (CKIs)
_
stuff happens
activity
cell cycle progression
transcription translation
12
A. Hershko
13
Two E3 enzymes establish alternation of CDK
activity necessary for DNA replication
APC/C
SCF
Skp1 Cdc53/Cullin F-box protein Complex
Anaphase Promoting Complex/ Cyclosome
G1 cyclins (Cln-Cdc28)
origins load
M cyclins (Clb-Cdc28)
Cdk inhibitor (Sic1)
S cyclins (Clb-Cdc28)
G1
CDK ON
CDK OFF
M
S
origins fire
G2
M/G1 switch
G1/S switch
origin re-loading blocked
14
CDK activity fires replication and then prevents
re-replication
Start (Cln-Cdc28)
CDK antagonists
Sic1
Sic1
APCCdh1
CDK (Clb-Cdc28)
CDK
P
P
P
P
P
P
Cdc6
Cdc6
Cdc6
MCMs
Cdc6
ORC
ORC
ORC
origins fire in S phase
origin loading blocked until end of mitosis
origin loading in G1 phase
Mitotic Exit Network (Cdc14 release)
15
Incomplete origin loading in G1 phase leads to
genome instability
Normal CDK inactivation
wild type
50 kb
origin
barrier
Incomplete CDK inactivation
sic1 deletion
improper replication/elevated DNA damage
chromosome loss
Etienne Schwob
16
How is Sic1 proteolysis "switched on" at START?
phosphorylation-dependent substrate recognition
P
17
SCFCdc4 targets Sic1, a Clb-Cdc28 kinase inhibitor
growth
scaffold
Cln3-Cdc28
adapter
E3
RING-H2
Cln1/2-Cdc28
recruitment (F-box)
Clb5/6-Cdc28
DNA replication
wild type
cdc4/34/53
cdc4/34/53 sic1

1 N
2 N
18
The CDK inhibitor Sic1 is degraded in late G1
phase (Start)
Sic1-GFP
19
A threshold of G1 cyclin (Cln)-Cdc28 activity
is needed to pass Start
SCFCdc4
DNA replication
Cln-Cdc28 activity
B. Futcher, F. Cross
20
Multiple phosphorylation sites on Sic1 contribute
to SCFCdc4 recognition
S76
T2/5
T33
T45
S69
S80
T173
S191
Sic1
MTPSTPPRSR
MTSPFNGLSSPQRSPFPK
MQGQKTPQK...PVTPSTTK
NNSPKNDA
PGTPSDKV
Cdc4 bound
30 input
J. Tang Q. Chen M. Mendenhall
multi-site mutants stable (Verma and Deshaies,
1997)
21
At least 6 of 9 phosphorylation sites are
required for efficient Sic1 recognition and
degradation in vivo
pSic1, 40
Cdc4 bound
Sic1
2p
5p
6p.1
3p
3p
6p.2
5p
2p
7p
6p.1
9p
0p
6p.2
7p
(sites re-introduced according to rank genetic
order)
9p
J. Tang
22
No obvious consensus in Cdc4 substrates (except
CDK sites S/T-P-X-K/R)
Cdc6
Sic1
T7
SAIPI
TP
TK
R
I
T2
/5
M
TP
S
TP
P
R
S
R
T23
DDAPA
TP
P
R
PL
T33
MQGQK
TP
Q
K
PS
T39
QFTDV
TP
ESSP
T45
NLVPV
TP
STT
K
S43
VTPES
SP
E
K
LQ
S69
NMGMT
SP
FNGL
T135
PLSLS
TP
R
S
K
D
S76
FNGLT
SP
Q
R
SP
S354
KRFLL
SP
T
R
GS
S80
TSPQR
SP
FP
K
S
T368
AQVPL
TP
TTSP
T173
KDVPG
TP
SD
K
V
S372
LTPTT
SP
V
KK
S
S191
NWNNN
SP
K
NDA
Far1
Gcn4
S87
SKPI
SP
PPSL
KK
T165
FLP
TP
VLED
13 others
4 others
23
How are multiply phosphorylated
substrates recognized by Cdc4?
Possible Models 1. Numerous low affinity
binding sites simultaneously engage multiple
phosphorylated residues 2. Phosphorylation
driven conformational change leads to exposure of
a cryptic binding epitope 3. Single binding
site interacts with multiple phosphorylated
residues (low affinity) in equilibrium
24
A cyclinE-pT380 peptide out-competes substrate
binding, whereas pSic1peptides do not
- cyclin E is stabilized in cdc4-1 cells -
SCFCdc4 ubiquitinates cyclin E in vitro -
Cdc4/Ago/Fbw7/SEL-10 in metazoans (Reed,
Elledge, Hariharan) - cancer-associated
mutations in hCDC4 4q32 region (Reed,
Haber, Hariharan)
J. Tang, P. Nash
25
A CycET380 phosphopeptide binds to a single class
of high affinity site on Cdc4
Fl -ASPLPSGLLpTPPQSGKKQS
Hill Coefficient .997
KD0.91 µM
log(Y/1-Y)
Fluorescence Polarization
log(Cdc4/Skp1)
Cdc4/Skp1 (mM)
P. Nash, S. Orlicky
26
A SPOTS peptide array based on CycET380 defines
the Cdc4 Phospho-Degron (CPD)
Cdc4 Phospho-Degron (CPD) I/L-I/L/P-pT-P-(RK)-(RK
)-(RK)-(RK) (Basic Residues Disfavored)
VS
CDK Consensus pS/T-P-X-K/R (Basic Residues
Preferred)
P. Nash, F. Gertler, S. Orlicky
probed with anti-Skp1/Cdc4
27
Absence of secondary structure in Sic1
Full length
ºC
1-90
N-terminus
1-120
ºC
S. Orlicky
28
Synthetic concatamers of low affinity CPD
sites bind Cdc4 with high affinity
GST-GLTSPQRSPFPKSSPPRS3,6,9,12
S76 S80 T5
GST-VTPSKPVTPSKPVTPSR3, 9,12
T45 basic residues
J. Tang, S. Orlicky
29
3 conserved essential arginines define a CPD
binding site
R534A
R443A
R467A
R485A
WT
Relative CPD Binding
S. Orlicky, J. Tang, A. Willems, P. Nash
30
Optimal CPD
Sub-Optimal CPD
versus
Cdc28
kinase
(xxPVTPxK/Rx)n
xxLLTPxxx
???
Sic1
Cdc4
Cdc4
Degradation
Degradation
But why bother with multiple sites?
31
A single optimal CPD site (LLTPP) confers Cdc4
binding and ubiquitination in the absence of all
other sites
T45cycE
LLT76PP
S76S
T45T
wt

-
-

-
-

-
-
Cdc4 bound
-
-
-






Cln-Cdc28
Sic1-Ubn
Sic1
Sic1CPD
Sic1T45,S76
Sic1WT
pSic1
J. Tang
32
.
and restores Sic1 instability in vivo.
return to 25ºC
25ºC
37ºC
LLT45PP
10
15
20
30
40
60 min
T45T
S76S
Sic1WT
LLT76PP
T45CycE
Sic1CPD
WT
cdc28-13 strains
GAL1-SIC1 ON (all in 9 site mutant)
J. Tang
33
However, Sic1CPD levels are heterogeneous in G1
phase cells
(both as GFP fusions)
Jean Tang
34
Which results in failure to restrain DNA
replication...
SIC1
SIC1CPD
J. Tang
35
and genome instability
SIC1
sic1D
SIC1CPD
chromosome loss rate
lt 1 in 105
100 X increase
(rampant)
J. Tang
36
So far.
The Cdc4 Phospho-Degron (CPD) I/L-I/L/P-pT-P-(rky)
-(rky)-(rky)-(rky) is at odds with the CDK
consensus
Multiple sub-optimal CPDs in Sic1 establish a
minimum CDK-free window needed for assembly of
replication origins and genome stability
But why multi-site phosphorylation? ? to set a
high threshold ? to precisely build a threshold
? ultrasensitivity
37
Recall Xenopus maturation
How do oocytes decide when to mature? How is this
decision made irreversible?
Maturation depends on a MAPK cascade
PG
Cdk
mos
Mek1/2
p42/44Erk1/2
MAPKKK
MAPKK
MAPK
only 30 fold ?
3 X109
4 X1011
1011
of molecules
38
A switch fantastic the MAPK cascade in Xenopus
maturation
population activity
the experiment 200 oocytes, each in a tube...
single oocyte activity
Jim Ferrell Science 280, 895 (1998)
39
Ultrasensitivity and bistability
PG
Cdk
mos
Mek1/2
p42/44Erk1/2
MAPKKK
MAPKK
MAPK
nH 42!
(cf. haemoglobin 2.8)
40
Routes to Ultrasensitivity
1. Multi-site phosphorylation 2. Saturation of
back-reaction 3. Feed-forward loops 4.
Stoichiometric inhibitors 5. Translocation
6. Zero-order ultrasensitivity
MAPKKK
MAPKK
MAPK
Amplification 20 to 100 fold Cooperativity nH
42
(J. Ferrell, see TiBS 21 460-466 for more details)
41
So for Sic1 degradation, multi-site
phosphorylation ? ultrasensitive response
Michaelian (9X response requires 81X stimulus)
response
Sic1 degradation
stimulus
Ultrasensitive
kinase
NB requires distributive phosphorylation
response
Sic1 ? Sic1-P1 ? Sic1-P2 ? Sic1-P3 ? Sic1-P4 ?
Sic1-P5 ? Sic1-P6...
stimulus
Nth order dependence on kinase
42
Multi-site phosphorylation of Sic1 may lead
to switch-like onset of DNA replication
Clb-Cdc28
Sic1
Cln-Cdc28
Cln-Cdc28 activity
43
BUT why 6? i.e., how does a linear increase in
site density result in an apparent step function
for binding???
pSic1, 40
Cdc4 bound
Sic1
2p
5p
6p.1
3p
3p
6p.2
5p
2p
7p
6p.1
9p
0p
6p.2
7p
(sites re-introduced according to rank genetic
order)
9p
J. Tang
44
A high local concentration may drive binding to
weak secondary sites
Deshaies and Ferrell, 2001
45
Mathematical modeling suggests that a linear
increase in site density can result in
cooperative binding
Assumptions 1. single interaction between ligand
(Sic1) and receptor (Cdc4) 2. n ligand sites on a
flexible linear molecule, same on and off rates
for each 3. imaginary boundary sphere of radius
r, mean square end-end distance 4. B bound, P
proximate, F free states for ligand 5. in
state P ligand moves on independent paths
constrained only by length 6. diffusion limited,
no entropic considerations
Properties of critical parameters 1. kon is
proportional to n 2. koff is independent of n 3.
kcap is diffusion limited, independent of n 4.
kesc varies inversely with n, of the form kesc
Ce-cn, i.e., an exponential
P. Klein
46
BUT a polyvalent ligand binds cooperatively to
single site alone (an avidity effect of
sorts)
bound
proximal
free
on rate gtgt
diffusion
P. Klein
47
More ultrasensitivity in the G1-S transition
Mechanisms (J. Ferrell D. Koshland)
M.E.N.
1. Multi-site phosphorylation Sic1
(distributive) 2. Saturation of back-reaction
Cdc4 binding blocks pase? 3. Feed-forward
Clb-28 on Sic1 Cln-28 4. Stoichiometric
inhibitor Sic1 Cln-28 on self other? 5.
Translocation Cdc14 back to nucleolus? 6.
Zero-order ultrasensitivity?
Cdc14?
Sic1-P1, P2 P6, P7, P8, P9
Sic1
SCFCdc4
Cln- Cdc28
Clb5/6- Cdc28
DNA replication
see TiBS 21 460-466 (1996)
48

• Ultrasensitivity in other responses...
• mitotic onset
• - Cdc2/Cdc25 positive feedback loop, multi-site
  phosphorylation
• mitotic exit
• - kinase/phosphatase network, multi-site
  phosphorylation
• critical cell size
• - multiple inputs/Cln-Cdc28 kinase threshold
• DNA damage
• - Rad9-Rad53 interaction (D. Durocher)
• mating response
• - MAPK cascade, multi-site phosphorylation,
  scaffolds
• voltage gated ion channels, lambda repressor.
  etc.

49
Metazoan SCF pathways (and multisite
phosphorylation?)