ubiquitination and DNA replication restart

DISS ETH NO. 15874 RttlOlp, provides a novel link between ubiquitination and DNA replication restart The cullin, A dissertation submitted to the S...
Author: Lora Holt
1 downloads 0 Views 8MB Size
DISS ETH NO. 15874

RttlOlp, provides a novel link between ubiquitination and DNA replication restart

The cullin,

A dissertation submitted to the

SWISS FEDERAL INSTITUTE OF TECHNOLOGY ZURICH for the

degree

of

Doctor of Natural Sciences

Presented

by

Brian Luke

Bsc. Honours,

Queen's University, Kingston, Ontario, Canada Born

May 28,1976

Citizen of Canada and

Accepted on

Belgium

the recommendation of

Prof. Matthias Peter- examiner Prof. Wilhelm Krek- co-examiner Dr.

Philippe

Pasero- co-examiner

2

Table of Contents

ZUSAMMENFASSUNG

5

SUMMARY

6

Opening

7

Remark

CHAPTER 1

-

8

GENERAL INTRODUCTION-1

Novel roles for cullin dependent ligases

8

Protein degradation by the Ubiquitin/Proteasome system

8 9

Ubiquitin and Ubiquitin Conjugation E3 Ligases DNA

replication of

11

19

and

repair replication origins

20

Regulation Replication slow zones Sensing DNA damage Checkpoints Repairing Double Stranded Breaks

22 24

-

The role of

ubiquitin in

DNA

25

28

replication and Repair and Replication

28

Ubiquitin conjugation/degradation Ubiquitin and Repair CHAPTER-II THE YEAST

30

CULLIN, RTT101P, IS INVOLVED IN GENOMIC 32

INTEGRITY AND DNA REPAIR

33

Introduction

36

Results rttlOl A cells have

a

checkpoint dependent G2/M delay

and accumulate with short

pre-anaphase 36

spindles rttlOlA cells exhibit

a

slower S

phase

than wild type cells and have fewer

firing RttlOlp is synthetically-lethal with the replication helicase Rrm3p overlapping functions rttlOlA cells are highly sensitive to the damaging agents MMS and sensitive to HU, and not sensitive to

an

firing replication origins 38

,

but

they

do not share

40 CPT but

only moderately

HO induced cut

phase checkpoint replication following MMS treatment but not HU treatment is impaired Cells lacking RttlOlp have increased DSBs and genomic instability

44 47

49

Discussion G2/M cell

cycle delay due to a defect during DNA-replication involved in the replication of proteinaceous chromatin regions not required for the Rad53p-checkpoint or DSB-repair, but specifically functions at

rttlOlA cells exhibit

a

RttlOlp is RttlOlp is collapsed replication forks RttlOlp is required to restart collapsed but not stalled replication forks The ability to restart collapsed replication forks is important to ensure genome stability Does the cullin RttlOlp function as an E3-ubiquitin ligase to restart collapsed replication forks?

Strains constructions and

manipulations

49 50 51 52 53 54

56

Material and Methods

DNA

42

is activated and executed in rttlOlA cells, however re-initiation of

The intra-S

genetic manipulations

56

56

3

56

Microscopy

57

with the Gal HA-RTT101 strains

Arrest/release

experiments analysis and BrdU detection 2D gel analysis and ERC isolation Dynamic molecular combing

57

PFGE

58 58

CHAPTER-III OTHER COMPONENTS OF THE RTT101P PATHWAY approaches to find other components of the RttlOl pathway Ubc4p and Ubc5p, interact genetically and physically with RttlOlp The N-terminal region of RttlOlp is essential for function indication a conserved role in adaptor binding A genomic wide screen to find other RttlO lp interactors Modification and regulation of the cullin RttlOlp

Using

60 60

Genetic

60

The redundant E2s,

substrate

CHAPTER-IV GENERAL DISCUSSION

cycle

71

71

contribution of RttlOlp

Checkpoints Checkpoint

65 69

71

Overview

The cell

64

and cell

cycle

71

arrest

inactivation to reveal

pathway replication initiation? DNA repair? a

involved in

Is the cullin,

RttlOlp, play direct role in So what is RttlO lp doing? Is RttlOl functioning as a ubiquitin ligase? Are there other components? Is RttlOlp regulated? Is this a conserved process? RttlOlp and the cancer connection Does RttlO lp

72

73 75 78 79 81 81 82 83

Remarks and Outlook

Concluding

72

CHAPTER- V CONTROL OF NUTRIENT-SENSITIVE TRANSCRIPTION

PROGRAMS BY THE UNCONVENTIONAL PREFOLDIN URI

84

Abstract

85

Introduction

85

Results

85

Discussion

90

Figures

92

References

97

98

SUPPORTING ONLINE MATERIAL

98

Materials and Methods

107

Supporting Figures

115

APPENDIX I

115

Materials and Methods Determination of DNA content DNA prep for 2D 2D gel

Preparation

by

FACS

REFERENCES

115 115

gel analysis

of BrdU-substituted DNA

PFGE

analysi s

116

plugs

117 117

118

4

ACKNOWLEDGMENTS

129

CURRICULUM VITAE

130

5

Zusammenfassung

Zuverlässige Rephkation

der DNS ist

essentieller

von

Bedeutung

für den

und somit für das

Überleben eines Organismus. Ist die Rephkation

beeinträchtigt,

können Fehler entstehen. Solche Mutationen führen

so

Instabilität und dadurch

zu

zu

genetischer

für Schäden, die

chemischen, mutagenen Substanzen, y-Strahlung, aber auch durch zelleigene

von

Rephkation

oxidative Metaboliten oder fehlerhafte

entstehen können. Eine Zelle

deshalb in der

Lage sein, solche Schäden reparieren

Verdoppelung

der DNS sicherzustellen.

Mittels der

Phase

Sprosshefe

Hefeprotein

dass das

um

der DNS

anfällig

Krankheiten wie Krebs. DNS ist

Zellzyklus

zu

lösen.

in sich

S.cerevisae als

der DNS

Replikationsgabeln

konnten wir

nicht

wiederhergestellt,

Doppelstrangbrüche reparieren;

Replizieren

von

beschleunigen (Michel

eine

aber

E3-Ligase Diese

doppelsträngigen

zu

können

von

Weiter haben wir bewiesen,

'checkpoints'

sog.

Wirkung

von

RttlOlp

et

für

Ubiquitin

oder

das

Hypothese

bei in sich

wird

von

E2-Ubiquitin Enzymen Ubc4p von

Ubiquitin-Ketten in

genetischen und

Cullin-Familie das Fortsetzen der

angehalten wurde, gewährleistet.

von

Zellextrakten anzutreffen. Dies

DNS-Replikation nachdem

eine

E3-Ligase

der

Replikationsgabel

Höchstwahrscheinlich existiert ein Protein, dass

Doppelstrangbrüchen ubiquitiniert hat.

Ubc5p

vitro

mit den obenerwähnten Daten, darauf hin, dass eine

Folge

und

ai, 2003a).

chromatinhaltigen Fraktionen

zusammen

zur

die

DNS wieder aufnehmen können.

gestützt. Zudem kann RttlOlp die Bildung

Synthese

RttlOlp benötigt wird,

reparieren, damit

führt dies oft

angehalten,

zwar

biochemischen Interaktionen mit den

weist,

zu

dass

RttlOlp-Aktivität aufweisen,

zusammengefallenen Replikationsgabeln.

ist in

konnten wir feststellen,

dies weist darauf hin, dass dieses Cullin keine direkte

RttlOlp ist höchstwahrscheinlich

RttlOlp

zeigen,

stabilisiert werden, auch ohne die

Kontrollpunkten

eine fehlerfreie

dazu dient, Probleme in der S-

generellen DNS-Reparaturmechanismen spielt.

Replikationsgabeln, die

um

werden kann. Werden solche defekte

fortgesetzt

Brüchen in der DNS. Hefen die keine

dass

RttlOlp,

zusammengefallene Replikationsgabeln

Rephkation

Rolle in

Modellorganismus,

der 'Cullin'-Familie,

Genaugenommen,

können,

zu

muss

wird und damit die Wiederaufnahme der DNS-

an

6

Summary Accurate DNA

organismal

replication

survival. If the

which may lead to

replication machinery

compromised,

and diseases like

genetic instability

attack from exogenous agents

is

including y-irradiation

errors can occur

DNA is vulnerable to

cancer.

and chemical mutagens,

however the accumulation of endogenous oxidative metabolites, and

occurring DNA replication with such

equipped to deal error

free

also result in DNA

errors can

replicative

and

and

cycle progression

is essential for cell

damage.

naturally

A cell must be

induced stress in order to

damage

ensure

genomic duplication.

Using

the

budding yeast,

is

protein, RttlOlp,

that the yeast cullin

problems. Specifically,

we

collapsed replication forks, deleted for

S. cerevisiae,

RttlOlp,

are

to

repair

is

Furthermore,

have demonstrated that

replication

we

efficiently

to

result in

a

a

we

have found

cope with S

phase

in the re-start of

double stranded break. Cells

double stranded break

intact and the cullin does

repair machinery

organism,

RttlOlp is involved

frequently

competent

fully

required

have shown that which

model

as a

directly

function

indicating

that

during repair.

stalled, but checkpoint stabilized

a

fork is able to restart in the absence of

RttlOlp.

As

well, RttlOlp, is

localized to the chromatin fraction of cell extracts.

RttlOlp collapsed forks,

is most as

it shows

ubiquitin conjugating RttlOlp

can

most

is

resumption

enzymes,

together

required

likely

genetic

as an

E3

ubiquitin conjugating

ligase

Ubc4p

at

a

for

these data

replication

and

Ubc5p.

It has been shown

provide

the first evidence that

fork re-start, and suggests that

double stranded break site, that becomes

of DNA

synthesis.

at

and biochemical interactions with the E2

previously

enhance the formation of ubiquitin chains in vitro (Michel

Taken

ligase

likely acting

a

a

et

al.

that

2003).

cullin based E3

target protein exists,

ubiquitinated to

allow the

7

Opening

Remark

The work that I have done which

are

thesis

was

during

my PhD thesis

very different in terms of

devoted to

determining

topic.

The

was

split

minority

the role of the novel

into two different

of

(approximately 1/4)

prefoldin protein, URI,

sensing pathway, through

transcription factor, Gcn4p.

This work is not introduced in the

the role of the cullin,

Chapter

RttlOlp,

to efficient and accurate cell

potential

cycle progression,

in the

in the

general introduction,

V. The remainder of my thesis work

and how this

of my

the translational control of the

activation of the amino acid

however it is covered in

projects,

E3

was

ligase component

budding yeast,

focused

on

contributes

S. cerevisiae.

8

1- General lntroduction-1

Chapter

Novel roles for cullin The

dependent ligases of proteins

regulated degradation

sequence

interrogate

RttlOlp.

the role of the novel and

We

is

only recently begun

has

mainly

to be elucidated.

budding yeast, Saccharomyces cerevisiae,

employed

about the role of

RttlOlp

ligases

homology it has long been known that other cullin proteins exist, their

cellular functions have We used the

cullin-based E3

(Cull) containing complex. Although

been addressed in the context of the Cullin 1

by

by

both

RttlOlp

previously

genetic

in the cell.

model system to

uncharacterized cullin

and biochemical

approaches

My studies have led

replication

involved in both DNA

intimately

as a

us to

to

protein,

try to learn

more

the conclusion that

and DNA

repair following

damage. Therefore, the following introduction will provide background information both

of cullin based

ubiquitin ligases, focusing

on

ligases.

I will also introduce

aspects of DNA replication and repair and finally the last part of the introduction will describe what is

currently know

about the role of

ubiquitination

with

regards

DNA

to

metabolism.

Protein

degradation by

The cell

activity, signals,

which

can

is driven

in turn

by

Ubiquitin/Proteasome system

oscillations of

cycle

until the

on, however

be shut off to

ensure

whereas STOP

that

signals

stage. The best way

irreversibility.

they

to

cyclin dependent kinase (CDK)

phosphorylate targets, thereby generating

which promote entry into the next cell

slow the cell

signals

cycle

the

previous stage

also need an

to

abolish

a

completed.

be turned off. For

event is not

need to be

is

cycle stage,

repeated

destroyed

signal

is to

to

both "GO"

and "STOP"

CDK

activity

example

twice within the

signals

can

a, GO same

signal, cell

allow entry into the next cell

destroy it, thereby ensuring

The controlled destruction of

protein signals

within

a

turn

cell is

which these must

cycle,

cycle

9

accomplished with great This system is based

on

accuracy and

the

principal

target, it becomes recognized

that when

that

degraded become polyubiquitinated, resulting spatial

ensure

addition to

and

temporal protein

degradation

of mis-folded

Ubiquitin

and

Ubiquitin,

76 amino acid

a

to

ubiquitin

in

an

ATP a

which is

al. 1984).

protein.

highly

Many regulatory proteins to

a

ubiquitin

chain. The

generically

conserved

are

cycle.

In

the

conjugation

to the

to an

glycine

et al.

active-site

responsible for recruiting an

a

thio-ester bond

residue

cysteine

on an

by

an

E-amino group

E2, ubiquitin

E3

1.

on

ubiquitin

or

can

the

conjugated to

is recruited to the 26S proteasome which

Figure

(G76) of

1981). Ubiquitin is then

rounds of this reaction the substrate

This reaction is outlined in

ubiquitin

E3, which

the substrate to the E2, which

accepting lysine

of

targeted substrate (Hershko

The last step of the reaction is carried out

repeated

protein

subject to rapid

termed El, E2 and

dependent manner (Ciechanover

trans-thioesterification

polyubiquitin chain, the

most

conjugate ubiquitin

subsequently transfer ubiquitin to After

is the

residue in the El and the C-terminal

conjugating enzyme.

.

cell

single

1998). The El, ubiquitin activating enzyme, forms

cysteine

transferred via

protein

a

three amino acids substitutions from yeast to human

activate and then

and Ciechanover

ligase,

protein,

the action of three enzymes,

a

distribution within

key

This system

"targeted degradation".

in

is

that should be

proteins

endocytosis, receptor down-regulation and

proteolysis following conjugation

between

those

Specificity

Ubiquitin Conjugation

ubiquitin (Ozkaynak et

cooperate

only

enzymatic

via

peptides

proteins.

eukaryotes, with only

requires

to a

playing a fundamental role in cell cycle progression, ubiquitination has

been shown to be involved in

among

ubiquitin chain is conjugated

a

substrate for the 26S proteasome, which

ubiquitin/proteasome system, ensuring

to

ubiquitin/proteasome system.

al. 1983; Hershko and Ciechanover 1998).

proteolysis (Hershko et

helps

the

the target into small amino acid

consequently degrades

to the

as a

efficiency by

targeted

a

enzymatically degrades

10

A. Ub-Ub-Ub

J

.

Substrate

B.

AMP+PPj

ATP

°

^A

9

2 ^» Ei~S-C-Ub—-"* E2-S-C-Ub

Ub-C-01O

Protein

Ë2-SH

ATP

ADP+Pj

Ub

Vl^

h

Ub-peptides-*

Protein—(Ub)n

26S Proteasome

Peptides

Ub

I8 Amino acids

Figure

1. A. A schematic of the

reactions

as

Hershko and

For

chain

poly-ubiquitin

a

for

and Rechsteiner

proteolysis.

1986), whereas

conjugates

more

et al.

1984b).

26S proteasome.

Rad6p/Ubc2p,

as

K48

Although

formation, they do

not

well

shown

to be

as

linkage

which

can

linkages

other

than

most

a

lysine residues

and

al. 1995), which has been shown to

Ubc5p

play

a

are

ubiquitin.

Hough

may result in

something

other than

mono-ubiquitinate

a

substrate but is

proteolysis (Hershko a

have

substrate with

an

within

targeted degradation.

Ubc4p

are

al. 1984b;

substrate with

likely

there

et

biochemically that

degraded

result in

manners as

ubiquitin chains primarily

1984a; Hershko

does not result in substrate

was

likely

et al.

different

a

(methyl ubiquitin),

1985). Furthermore it

(Hershko

be formed in many different

degradation (Hershko

Me-Ub

unable to form

is

can

it has become clear that K48-G76 linked

protein

conjugate

implicated in the ubiquitin system. B. The conjugation of ubiquitin to a substrate, from

within the 76 amino acids that make up

potential acceptor lysines

example,

mark

enzymes

Ciechanover, 1998.

The seven

generic

described in the text of the activation and

a

short

increased

affinity

for the

used in chain

example, the

involved in K63

long ubiquitin

conjugate

ubiquitin are

For

a

and Heller

E2 enzyme

linkages (Spence

fundamental role in DNA

repair (see

et

11

When K63R

below). are

dependent

the E2s

on

the structure of the

Ubc4p

expressed

1987)

and

ubiquitin chain

differentially placed lysines Kumar et al.

is

repair defective. K29

rendered DNA

also

ubiquitin

as can

seen

linked

only

of

source

conjugates

Ubc5p (Spence

drastically

in

Figure

are

et al.

ubiquitin,

different

the cells

also formed and this is

1995). It is

most

conjugate

as

structures

likely the

(Vijay-

2.

KW

K«J

Figure

the

that determines the fate of the substrate

leads to

be

as

Ubiquitin. The center and right images are two right turns about the y-axis. sticking up towards the top of the page and the chain forming lysines are accordingly. Taken from, Ubiquitin, and the biology of the cell, chapter 2,1998.

2. The structure of

The C-terminal G76 is

marked

There is

activating

only

one

El

the total cellular

protein

pool

of

in both yeast and

ubiquitin.

E2s, 13 in yeast (Hochstrasser 1996), and hundreds. It is the E3 substrate will be

E3

which

provides

However there

far

an

the

are

there

are

many

unknown number of E3s,

specificity

more

possibly

and determines which

degraded.

Ligases E3

trying

to

into the

ligases

identify

come

can

multi-subunit

in various forms, which has

provided

the

them. The fundamental function of the E3 is to

proximity

substrate. E3s a

ligase,

so

human, which is responsible for

of the E2 enzyme be

composed

complex.

of

a

Some E3s

thereby promoting ubiquitin single protein, however,

participate directly

by transiently accepting ubiquitin before

it is

in the

conjugated to

can

major hurdle bring

in

the substrate

transfer to the also be made up of

ubiquitination reaction

the substrate, whereas

12

others

form

never

E3-Ub intermediate. The different types of E3s will be

an

described below with

emphasis

on

the SCF and its role in cell

cycle integrity.

HECT domain HECT domain

C-terminal

on a

This is

containing

lysine

dependent

on

E3

proteins

are

unique

within the conserved HECT domain

the action of

an

E2

they accept ubiquitin

in that

(Scheffner

et al.

1995).

enzyme, which

ubiquitin conjugating

promotes the formation of a Ub-HECT intermediate. Ubiquitin is subsequently the substrate from the E3, rather than from the E2

passed onto

genome encodes for

five

(Huibregtse

et

a

al.

large family of HECT

domain

1995). HECT proteins

can

proteins,

bind

The human

directly.

whereas yeast has

directly

to

the

only

targeted

substrate; for example the yeast Rsp5p HECT containing protein interacts with the

large

subunit of RNA

destruction

polymerase after

(Huibregtse

substrate via

an

p53 through

the

et al.

1997).

as

which marks it for

damage

However

adaptor intermediate,

adaptor E6-AP,

DNA

is the

and results in

some

case

HECT

proteins

with human

proteasomal

interact with the

E6, which binds

to

al.

p53 degradation (Scheffner et

1993). In contrast to

ubiquitination via

ubiquitin conjugation permease,

Gaplp,

is

does not

always

ubiquitinated

media conditions. This

in

the N-end rule

lead

proteolysis.

et al.

HECT

The yeast

dependent

general

Rsp5p dependent manner when

an

ubiquitination results

endocytosis pathway (Hein

pathway,

in the internalization of

amino acid

grown in rich

Gaplp

via the

1995).

N-end rule E3s The human E3a based

on

protein's to

protein

their N-terminal amino acid

and the yeast

with the E2,

and then becomes

Ubc2/Rad6p,

ubiquitinated

or

bulky

and

and

hydrophobic

degraded.

it

a

can

bind

This E3 also interacts

which is also conserved from yeast to

1989;Dohmenetal. 1991).

substrates

(Hershko and Ciechanover 1992). When

N-terminal amino acid is either basic

E3a/Ubrlp,

Ubrlp, recognize

man

(Reiss

et

al.

13

APC/C

-

Anaphase promoting complex/Cyclosome

The APC

B-type cyclins

discovered

was

immunoprecipitations APC is

sequence domains bear

a

in

screens

Xenopus

(Zachariae

resemblance to

are

dependent

manner.

The mitotic

Pdslp/securin contain 1996; Ciosk

et al.

them at the

cohesin

complex

anaphase

onset,

After correct

dependent

and

the APC via

It

al. 1995;

of

found

was

King

et

al.

1995). The

striking primary

no

subunits, Apc2p and

two

Rbxlp,

can no

a

which

are core

components

conserved nine amino-acid motif

mutagenesis

or

of

a

substrate's destruction

longer be ubiquitinated

B-type cyclins

destruction boxes and

as

are

well

the

as

in

an

APC

inhibitor

anaphase

targets of the APC (Cohen-Fix

et al.

1998).

Once cells have

align

it

1995).

stability

Nasmyth 1999).

called the destruction box. The deletion as

the

regulate

many of which have

Cullp/Cdc53p

recognized by

box results in its stabilization,

et

1998). However

of the SCF (see below) (Zachariae and Substrates

(Irniger

extracts

al.

et

et al.

to

in yeast, and biochemical

subunits,

of at least 12

composed

Apcl lp,

(King

upon the exit from mitosis

independently through genetic

ability

account of its

on

replicated

their DNA

metaphase plate. in order to

manner

condense their chromosomes and

At this time chromosomes

ensure

correct

thereby minimizing

spindle

they

bipolar spindle

are

the chances of chromosome

degrades Pdslp/Securin (Visintin

together by the

attachment before

mis-segregation.

is established the APC becomes activated in

and

held

et

a

Cdc20p

al. 1997;

Harper et

al.

2002). The destruction of Pdslp/Securin results in the release of Esplp/separase, which cleaves the

(Uhlmann

et

al.

Scclp

subunit of

onset

1999) (Figure 3). The mitotic cyclin, Clb2p, is degraded

the exit from mitosis while at the

origins

cohesin, thereby allowing anaphase

for the next round of S

same

time

allowing

phase (see below).

accessory/activating factor, Cdhlp,

rather than

the

This is

to stimulate

licencing of replication

accomplished by

Cdc20p (Visintin

et al.

the

1997).

14

Inartne

c>f«)|»!asm

nucleus

Securin

^^

Following phosphorylation ot Pdslp/Securin by Cdk I/Cdc28p it is degraded by the APCrill2
»

r

-Ifv

-

Mj

il

i

1

S (lit I

""M*

rtl

\

1

I

„i ,

t

Il

f

Ni.»,.--"

$

Pi

Ö

0

Ahhii«

il

»!

*

'"I'

1

l'if.

f»I

l"

-^11

'

V>'

>

'iMt'X

Figure 12. Taken from Lisby et al, 2004. choreography of DSB protein recruitment. A-B MRX complex, Tellp and Sae2 are recruited, followed by C MRX dependent strand resection, the ssDNA is coated by D RP-A which recruits Ddc2/Mecl, Rad24/Rfc2-5, and Ddcl/Radl7/Mec3. E-F. Once cells have entered S phase, RadSl and Rad52 are recnnted, which promotes sister chromatid strand

invasion.

recruited to

Hydroxyurea (HU)

stabilized due to the activation of the

Rad53p checkpoint.

Interestingly, Rad51p stalled forks, which However Rad5

(eg.

30

lp

are

and

and

Rad52p,

maintain

a

and

a

rad53A

recombination,

as

activity

it promotes 5'

prolonged checkpoint activation. is

up-regulated

in

fork. The

-

is

early

required at

of

et

al.

2004).

HU exposure

they

also

inability

to

well

as

proteins

collapse.

for efficient

as

It has

recently

homologous

the break site, which

In the absence of CDKl

compensation (Ira

due to the

response

independent

3' resection

prolonged

be maintained, moreover,

background again

form at HU stalled forks

been demonstrated that CDKl

not

longer

can no

stably paused replication Mec3p,

are

foci do form in response to

hours), when fork stability

form in the presence of HU in

Ddclp

Rad52p

activity

ensures

the NHEJ

pathway

28

When the NHEJ well

as

the Mrel

pathway

Ip/Rad50p/Xrs2p complex

stranded break. After strand

lp complex,

Mrel

activated, the Ku proteins, Ku70p and Ku80p,

is

are

targeted

as

the site of the double

to

resection, perhaps via the endonuclease activity of the

the DSB ends

are

ligated in

a

unspecific

sequence

manner

(van

Gent et al. 2001).

The role of

As discussed above, mono

and

in DNA

ubiquitin

ubiquitination

poly-ubiquitination,

linkage. Recently

in all forms

a

regards

to

fundamental role in S

replication

and

phosphorylated by CDKl,

have

already

it is

ubiquitin

chain

and DNA

ubiquitin

and

including variable

linkage is also

ubiquitination conjugation In the

repair.

following

modifications

ubiquitin-like

repair.

pre-RC components

and further re-activation of

Repair

be detected in many forms

phase

and

Ubiquitin conjugation/degradation The

and

it has been shown that

section I will outline what is known about

with

can

and the type of

from K48 to K63

plays

replication

a

Replication

complex regulation

fired

origins

of

to

replication.

prevent the re-loading

Cdc6p

Once

is

targeted for ubiquitination and degradation by the

SCFCdc4 E3 ligase, and then further prevented from pre-RC formation by interaction with Clb2, which is then itself

1995; Drury to

be

et al.

regulated

in

following origin

degraded

a

very similar way to

initiation

and

Cdc6p

degradation by

the APC/C

et al.

licencing

2003; Hu

et

al.

(Piatti

et

al.

is

ubiquitinated

by

(Figure 13).

DDB1

degraded

(Feng

and

Geminin, which is subsequently

the APC/C

Yoshida et al.

and

2004). As Cdtlp levels rise, it remains

via interaction with

Lygerou 2002;

in yeast

Cdc6p. Cdtlp

by the Cul-4A based ligase containing

upon the exit from mitosis

Nishitani and

for

1997). In human cells the licencing factor, Cdtlp, has been shown

Kipreos 2003; Zhong inhibited from

targeted

(Wohlschlegel

2004), in

an

et

al. 2000;

analogous fashion

to

Clb2

29

AK/C

[M'i

/A t D11

f

f

""*!i^^?J**^^^!^""^"^"'"'"^*

WK*

S [I've

G

Mi'i't

CDK*

a? rflitf>-

i

Figure 13 Taken from Li and Blow, 2003 At the end of mitosis the inhibitor of Cdtlp, Geminin, degraded by the APC/C leaving Cdtlp competent for the licencing reaction Once a licenced ongin has fired, Cdtlp, is degraded in a Cul-4 dependent manner.

In contrast the S. via

cerevisiae, Cdtlp is stable throughout the cell cycle and is regulated

sub-cellular localization rather than

fact that the cells

is

regulates

these

degradation (Tanaka

replication

events

and

Diffley 2002).

The

through protein degradation

augments the importance of irreversibility of these reactions within the cell cycle, thus

ensuring

that

licencing

As cells enter S

the

production

of

factors do not become competent for

phase

one

of the basic

deoxyribonucleotides (dNTPs)

a

reduction reaction

,

are

regulated

in

a

cell

degradation, whereby of S

is

Rnrlp

Spdlp (Sp)/Smllp(Sc) (Woollard

et

proteins

are

to

initiate

incorporated

negatively regulated by

manner

via

an

replication,

into the

converts NTPs

new

interacting protein and

specifically targeted et

for destruction

at

the onset

al. 2001 ; Bondar et al.

2004). In S. pombe, the Cul4p/Pcu4p and the substrate specific adaptor Ddbl et

al. 2003; Bondar et al.

cerevisiae, the E3 ligase has not yet been identified. So far has not yet been identified in the human genome.

Smllp

proteasomal dependent

phase, thereby de-repressmg Rnrlp activity (Zhao

required for Spdlp degradation (Liu

is

into dNTPs

1999). Spdlp

al. 1996; Chabes et al.

cycle dependent the

to be

protein, Rnrlp,

strand. The ribonucleotide reductase

through

requirements

re-licencing.

a

2004), whereas

protein

are

in

S.

inhibitor of RNR

30

and

Ubiquitin The

Repair cell nuclear

proliferating

in yeast. PCNA, which forms RNA-DNA

primed

trimeric

a

by

the POL30 gene

around DNA, is loaded onto Pola

clamp

clamp loader, replication factor

the

primers by

is encoded

antigen (PCNA),

C

(RF-C)

(Tsurimoto and Stillman 1989; Tsurimoto and Stillman 1991). PCNA interacts with Polo and promotes

directly

translationally

modified

(K164) (Hoege

et al.

replicative processivity.

both SUMO and

by

as

sizlA cells, which

the

are

for

impaired

sumo

hypersensitive

damaging agent, indicating

independent

phase.

the was

same

lysine

ligase

function

to DNA

that K164 is

a

important

required

replication defect,

(Hoege

damage

residue

shown that PCNA

However sumolation is not

K164R mutated PCNA does not have

contrast K164R cells are or

on

2002). Using synchronized cells it

becomes sumolated at the onset of S PCNA function

ubiquitin

PCNA is post-

et

al.

UV

by

damage repair

moving replication

fork encounters

damage along

the

template

strand, the E2 enzyme, Rad6p, along with the ring finger protein, Radl 8p, add

single ubiquitin

to the

to

as was

and

trans-lesion

Ubcl3p

replicate past

then add

PCNA is not

PCNA there appears to be

Figure

14.

does not affect the

polymerase, polr^ (Kannouche

a

promote template switching

ubiquitinated

mono-ubiquitination

demonstrated in cultured human cells, promotes

which is necessary to

Mms2p

The

onto PCNA.

PCNA however,

2004),

light

of sumolation.

When the

switch,

do

In

2002).

induced either for

nor

for

the

K63-1 inked

to

the

area.

2004; Vidal

competition

poly-ubiquitin chain,

which is

polymerase et

al.

ubiquitination

predicted

The multi-

this process. Therefore in the

between

stability of

The E2 enzymes

undamaged sister chromatid.

degraded during a

damaged

et al.

a

a

case

of

and sumolation.

31

K127 K164

Ubiquit.nat.on

•AW/*/

Multi-K63 linked

Mono

K,164

,v#f Lnor-free DNA repair

F

x

%^

PCNA

SUMO modification

Figure 14. SUMO and ubiquitin compete for the same lysine During replication PCNA is ubiqumated on K164, however becomes ubiquitinated and then further poly-ubiquinated after damage induction.

Other targets

including

are

also

degraded during

the above mentioned

and enhance

repair. Cdtlp

DNA

Spdlp (Bondar

is also

degraded

et

damage

expressing

after exposure to UV irradiation

likely

to

are

not as

MMS

to DNA

(Chang

et

damage.

one or

multiple targets

al.

as

ligase components

In addition,

the

single

most

RttlOlp, is

likely

also

al. 2002), and further it has been shown to possess

ubiquitin conjugation activity (Michel

has

et

factors has been

sensitive to MMS

ubc!3A cells, indication that these

targets in response

hypersensitive in vitro

or

(Hu

targets probably exist. For example, cells

more

the PCNA K164R mutation

radô, mms2A, have other

it is clear that

stress

al. 2004), to increase dNTP levels

2004). Although the regulation of many repair and replication

extensively studied,

replication

or

DNA

in the DNA

et

al. 2003),

damage

revealing

response.

that it also most

32

Chapter-ll The yeast cullin, RttlOlp, genomic integrity and DNA repair The work in this

chapter

was

is involved in

submitted for publication

RttlOlp, is required for recovery after damage and replication stress

The S. cerevisiae cullin, DNA

Brian

Luke1,

Lionel

1

Gwennaelle

Versini2,

Malika

Jaquenoud1'3, Thimo Kurz1,

Pintard1, Philippe Wiget1, Philippe Pasero2&,

and Matthias

Peter1*

Swiss Federal Institure of Technology Zurich (ETH)

Biochemistry Hoenggerberg HPM

Institute of ETH

G 10.0

8093 Zurich Switzerland Tel: 441 1 632 31334 / Fax: +41 1 632 1269

2

Institute of Human Genetics, CNRS

141 Rue de la Cardonille

34396

Montpellier

France 3

Present address:

Department of Biochemistry University of Fribourg Switzerland &

To whom

correspondence

should be addressed:

[email protected] [email protected]

Running Title: The S. cerevisiae cullin, RttlOlp, instability following replication collapse

is

required

to

prevent genomic

Key words: ubiquitin, E3-ligase, Cullin, DNA damage, RttlOl, Cul8

33

Introduction Accurate

replication

complex signaling high fidelity

and

network that

precision.

cycle progression

and maintenance of

that DNA

ensures

organism's genome requires and

duplication

onset of

DNA-segregation during

a

repair occur with orderly

These processes must be coordinated with

prevent the

to

an

cell

mitosis before

DNA-replication and repair are completed. stress, cells activate

Upon genotoxic arrest

the cell

activation of

cycle

until the lesion is removed

DNA-checkpoints

in either S

metaphase/anaphase

transition

2001 ;

2003). Some of the

Agarwal

et

al.

by

replication/damage checkpoints kinases

Meclp/ATR

kinases

are

and

and

responsible

or

or

repaired.

most

Rad53p/Chk2 (Gardner

an

arrest at

Pdslp (Clarke

the

et al.

upstream components of the

partially overlapping PI3-kinase-related Rouse and Jackson

phosphorylation of two

et al.

or

budding yeast,

G2 phase results in

Tellp/ATM (Carr 2002;

for the

In

stabilization of the securin

include the

to slow down

checkpoint pathways

1999; Sanchez

et

2002). These

downstream kinases,

al. 1999), which in

turn

Chklp activate

targets involved in DNA-metabolism and repair. Chklp and Rad53p have also been shown to

directly phosphorylate Pdslp,

which prevents its

Anaphase Promoting Complex0*20 (APCCdc20) (Gardner et 1999; Wang

et

(Agarwal

et al.

pathways

can

al.

2001),

most

likely by inhibiting

degradation by

al. 1999; Sanchez et al.

the interaction with

2003). Mutations that affect faithful DNA-replication

lead to

genomic instability,

as seen

the

Cdc20p

or

checkpoint

in many human diseases

including

cancer.

Genomic

integrity

checkpoint signaling

is not

and DNA

only dependent

replication

modification of these enzymes to the

the APC

or

cullin-based

checkpoint pathways.

complex via

that

right place. Regulated ubiquitinylation

triggered by and

ensure

an

Cullins

characterized based

adaptors (Pintard

on

et al.

and

they

and

provide

a

the enzymes

repair, are

but also

requires

activated at the

control many cell

scaffolding

covalent

right

time and

of substrates

cycle

transitions

function within the E3-

ubiquitin-conjugating

Pagano 2004).

directly involved in

subsequent degradation

E3-ligases

and interact both with the E2

adaptor (Cardozo

and

on

enzyme and the substrate

Different cullin families have been

their N-terminal interactions with different substrate

specific

2004). The budding yeast genome encodes for three cullin

34

proteins, Cullp/Cdc53, Cul3p Cullp

and its role in the SCF

and

Rttl01p/Cul8p (RttlOlp

from here on). Whereas have been

(Skpl,Çullin, F-Box) complex

studied, little is known about the role and subunits of the other

two

extensively

cullin

proteins

in

S. cerevisiae.

RttlOlp

can

support ubiquitination in vitro, and like other cullins, interacts via

its C-terminal domain with the

2003). Cells deleted for RTT101

with

preferential

are

viable, but display

of the retrotransposon

transposition rate

insertion

occurring

Ty 1 (rö

UV

damage (Chang

delay

in

the cullin

gene RAD9

RttlOlp

replication

or

In the

are

which is et al.

(Michel

is involved in

400

CPT and IR, but

al. 2003).

Finally,

are

2001).

not sensitive to

rttlOlA cells show

aspect of DNA metabolism,

most

a

DNA

likely

involves

DNA

replication

replication origins (Raghuraman

et al.

2001; Wyrick

are

timing

likely replicative helicase, mitosis and

the

distributed all

along

the

et

al.

2001),

replicative

the chromosomes

Inactivation of cyclin-dependent kinase

promotes the

licensing of replication origins

and

While the

is bound to

allows the

efficiency during

regions.

replication (Noton

recognition complex (ORC) activity

and

physically

at the end of mitosis

the next round of DNA

kinase

et al.

suppressed by deleting the intra-S-phase

S. cerevisiae,

Replication origins

(Cdklp) activity

phase

(Scholes

complete

budding yeast

from the telomeric to the centromeric

couples

transposition)

2003). Taken together, these results indicate that

some

activated with different

low CDKl

et

of Xy 1

regulation

repair.

approximately

process.

2002; Michel

anaphase progression,

checkpoint

which

et al.

60-fold increase in the

deletions has revealed that ml 01A cells

genotoxic agents MMS, HU,

sensitive to the

=

a

in front of tRNA genes

High-throughput screening of viable yeast are

et al.

ring-finger containing protein, Rbxlp (Michel

assembly

Diffley 2000).

origin

replication origins throughout the of the

cell

cycle,

licensing factors: Cdc6p, Ctdlp

MCM-complex (Diffley 2004).

followed

by

and the

This mechanism

DNA-replication. Subsequent reactivation of CDKl

Cdc7p/Dbf4p triggers origin firing

DNA

and the S-

replication (Lei

al. 1997; Zou and Stillman 1998).

Recently through specific

it has become clear that chromosomal

regions

timely

and accurate

of the genome

DNA-replication

including telomeric,

subtelomeric, centromeric, tRNA genes, rDNA and silenced regions requires

specialized factors (Ivessa

et

al.

2002;

Ivessa et al.

for

2003). These non-nucleosomal

et

35

form stable

proteinaceous regions

sites. For

replication pause

DNA-protein complexes, resulting

example,

in natural

Rrm3p (Bessler et

the Pif-like DNA helicase,

al.

2001), is only required for the efficient replication of these genomic regions while the bulk of

replication

localize to these

is

independent

al. 2000; Ivessa et al.

et

replication pausing

at these

significant pausing

was

survival of rrm3A cells

specific

found

sites is

throughout

depends

stabilize the

to

al. 2002).

et

combined with many as

on an

dramatically increased,

well

as

and

replication (eg. cdc45ts)

when

Rrm3p

replication machinery of

regions

known to be

an

mutants

order to

in rrm3A

cells, these cells require that

are

a

background

DSB upon

al., 2004), suggesting chromosomal

we

a

regions.

specific replication

found that rttlOlA is

(eg.

stalled forks

collapse (Ivessa

occur

with

a

et al.

higher

able to restart in

role of this

In this

putative E3-ligase

study,

we

stress or

DNA

checkpoint

arrest in

a

is

promoting

required restart

damage

pause sites.

or

damage,

timely

an

increased

rttlOlA cells

manner.

for accurate and

to

cell

cycle

with rrm3A

(Tong

et

replicate specialized

demonstrate that rttlOlA cells indeed exhibit

defect that results in

replication

ligases during

synthetic-lethal

breaks and gross chromosomal rearrangements.

DNA

as

replication fork is

interested in the function of cullin-based

progression. Recently,

RttlOlp

a

forks

complete replication.

We

a

replication

The double-stranded break

2000; Cha and Kleckner 2002). Since DSBs and fork stalls

frequency

(eg. radSOA,

compromised to

cannot be

rrm3A

unstable and often result in

highly

is absent. In

2004). These results suggest that

high protein density.

(DSB) repair machinery becomes essential in are

repair

et al.

in the absence of Rrm3p the traverse

2004; Torres

the

inviable when

are

with mutants involved in re-start of stalled et al.

no

phase checkpoint, which is

intact intra-S

paused replication complex

while

Interestingly,

the rest of the genome.

top3A, sgslA, srs2A) (Tong

successfully

found to

was

2002). In the absence of Rrm3p,

addition, synthetic-lethal analysis revealed that rrm3A cells

mrel 1 A),

Rrm3p

proteinaceous regions by chromatin immunoprecipitation (CHIP)

experiments (Ivessa

thought

of Rrm3p (Ivessa

Taken

DNA

of replication forks that have

unable to

were

our

data

have been unable to be maintained in

a

a

recover

most

likely by

result of

stable

from the

strongly suggest that

replication,

collapsed as

of double stranded

upon induction of

Importantly,

together,

complete

frequency

manner

encountering at

natural

36

Results

checkpoint dependent G2/M delay and accumulate with short pre-anaphase spindles rtt101A cells have a

To

investigate

progression al.

2003),

of

wild-type

the bud neck,

at

RttlOlp,

compared

we

a

short

spindle

during

et

mitosis with the nucleus

unseparated

and

cycle

cell

(Michel

and rttlOlA cells. Consistent with recent results

found that rttlOlA cells accumulate

we

positioned

the function of the yeast cullin

sister chromatids (data

not

shown). Importantly, time-lapse movies of single wild-type and rttlOlA cells

expressing

tubulin-GFP revealed that the

severely delayed

as

spindle elongation

onset

of

anaphase

in rttlOlA cells is

compared

starts at 120'

to 10 minutes in wild-

type cells (Fig. 1A and supplementary movie). We counted time 0'

pre-anaphase spindle

the short

until

elongation.

rttlOlA cells

The average

(n=5) and 7 minutes

was

migrated

half way into the bud

quite significant.

properly attached mitotic

moved to the mother/bud neck and measured the time

elongation time

variation

spindle,

in wt cells

The short on

subsequent

anaphase transition,

events

spindle

the cell-to-cell

dynamic

in rttlOlA cells is very

suggesting

including cytokinesis,

that the

cell

spindle

which results in chromosomal a severe

is

separation and exit

damage.

delay

and

elongation of the

that the cells do not

timing, indicating

at the

undergo

Taken

an

together,

metaphase

to

transition.

To determine when in the cell

functional

HA-Rttl01p

(Figure 1C).

Cells

hour. The culture

expression

of

were

was

protein.

grown in 2%

immunoblotting

in the

was

galactose

divided, and 2% glucose After three hours,

was

followed by

of marker

were

required,

expressed fully

and arrested in Gl with ct-factor for 1 was

added to

HA-Rttl01p

released

that

half to inhibit the

was no

by washing

was not

one

longer detectable

HA-Rttl01p is

morphological criteria,

proteins. RttlOlp

we

regulatable GAL/,70-promoter

glucose culture, indicating

The two cultures

cycle progression

cycle RttlOlp

in rttlOlA cells from the

HA-Rttl01p.

by immunoblotting unstable

100 minutes in

(n=4) (Fig. IB), although

these results suggest that rttlOlA cells exhibit

anaphase

approximately

several occasions,

from mitosis occurred with normal aberrant

was

to chromosomes and the cortex. After the sudden

the

the time when

as

out

FACS

required

a

rather

a-factor, and cell

analysis

and

for bud emergence

or

37

the onset of the bulk part of DNA

lacking RttlOlp of

Pdslp-myc

of

accumulated at the

and

RttlOlp-activity

or

Clb2p

in

Pdslp-myc

demonstrating

that

a

not

shown). However, cells

anaphase-metaphase transition,

significantly delayed.

was

is necessary

RttlOlp results

Clb2p

replication (data

during

every S-

specific delay prior

or

is not

required

G2-phase of the cell cycle, and lack

to the onset of

to

degradation

These results suggest that

anaphase. Interestingly,

accumulated in the Gl-arrested cells after

RttlOlp

and

depletion

maintain APC-Cdhl

of

RttlOlp,

activity.

r*V*{t-1

JiUJ^

Fig.

Figure

1.

rttlOlA cells have

a

Meclp dependent

G2/M

no

1

delay

pre-anaphase spindle. Tubulin-GFP (A) delay was used to visualize the spindle via live cell fluorescence imaging (supplementary figure (SI)). We noticed that the spindle was delayed in the onset of elongation from the time it reached the bud-neck (time 0) until the length of the spindle spanned the length of the two cells, although the timing was quite variable from cell to cell (B). (C) Cells with a Gal regulatable HA-RttlOlp were synchronized in Gl with alpha factor and released into media containing either glucose or galactose. The degradation of Pdslp-myc and Clb2p were delayed in the glucose culture where RttlOlp was shut off. (D) Logarithmic DAPI stained cells were scored for bud-neck localization (butterfly nuclei) As expected rttlOlA showed a G2/M delay that was suppressed by deleting the checkpoint protein Mec lp but not rttJOJA cells have

a

G2/M

and

accumulate with

a

short

Mad2p.

I

38

Degradation

of

the APC is inhibited

Pdslp by

in

DNA-damage checkpoint

unreplicated

checkpoint

mitotic

Mad2-dependent

of

case

by either the ATR/Mecl-dependent

in

case

of

activation of

of these

one

the mitotic index of

at

the mother/bud neck,

this G2/M

delay

was

compared

phase checkpoint kinase,

Mad2p

had

no

rttWIA meclA

effect

on

8% in

to

to

wild-type

the

cells

S

Rad9p

also suppress the mitotic

can

the ATR/Mecl S

spindle assembly checkpoint

slower S

phase fewer firing replication origins firing To determine if kinetics of

a

RttlOlp indeed

DNA-replication

Herpes simplex thymidine

cycle

2002), The cells

in

wild-type

and ml01A

kinase to allow BrdU

were

arrested in Gl

pulse-field gel electrophoresis (PFGE), chromosomes entered the

and visualized

by

approximately 45 rttlOlA cells, 50% BrdU

as

an

gels.

than wild

during S-phase,

functions

by

in the presence of BrdU. Genomic DNA

It has

delay

inhibitor

previously

of ml01A occurs

so

that

type cells and have

we

first

incorporation

2A.

was

then

prepared

and

migrated

via

only completely replicated were

Interestingly,

after 45 minutes, BrdU

DNA-replication.

and

a-factor and released into the cell

transferred onto was

minutes after release from the Gl-arrest in both

Figure

the

strains, which express the

antibody against BrdU. BrdU detection

shown in

compared

incorporation (Lengronne

The chromosomes

significantly delayed (lower graph). efficient

to measure

phase.

rtt101Acells exhibit

Schwob

caused

(Fig. 1D). Importantly,

cells, suggesting that the G2/M delay may result from DNA damage that

during

the

rttlOlA macïlA cells.

the mitotic accumulation of ml 01A cells.

been shown that deletion of

was

staining

by inactivating

wild type levels

deleting

while

or

by

unattached

or

of rttlOlA cells

used DAPI

or

23% of ml 01A cells accumulated with their DNA

delay,

suppressed

we

DNA-damage,

misfunction

delay

checkpoint pathways,

wild-type, rttlOlA,

Consistent with the mitotic

or

spindle

chromosomes. To determine whether the mitotic

by

DNA

while

a

membrane

possible wild-type and

wild-type cells showed

incorporation in ml01A cells

These data suggest that

was

RttlOlp is required

for

39

rttSOI

v»t

D

.....

£$.#.!$

*;

6*&

—1,

Jm^jia.ii

É

WO

«I*

0 30 .5 80 73 90 0 ÙO^SâOTS Su

(mh)

ffljO'

Wt

OaOJS 957590

0 30« 80 75 SOI m In)

D

wt

M ,2 *» ,* 40 9.

7

I*

•),

rttfSI /ttftf*

21 7«i

OS Stfe

I*

l+M4 01

0

,?

fik'f

2. S

Figure

Wild

BrdU.

PFGE

wild type bromide

1b2

jrtj

n

^0

di^lincü

^4

_ft

kb

Fig.

is slow and less

phase

origins

type and rttlOlA cells expressing TK

analysis

revealed that rttlOlA cells chromosomes

replicated staining and (upper) as

10 *>)

were

are

were

were

2

active in rttlOlA cells arrested

delayed

m

Gl and released in the presence of

completion of S phase compared gel (A) as judged by efhidium

in

slower to re-enter the

to

incorporation (lower). (B) Cells released from Gl into BrdU

and BrdU

were prepared and aligned linearly onto coverslips, followed by staining with anti-BrdU (green) and anti-DNA (red). Measurements between origins (IOD) and replication track length were made (upper panel). (C) The average IOD in rttlOlA cells was greater (68.8 kb) than the corresponding wt (54.2 kb). (D) BrdU track length is 21.7 kb m rttlOlA cells

were

arrested with HU and chromosomes

compared

The slower S

required

corresponding wild type.

to 14.2 kb in the

phase progression

for efficient initiation

between these two

or

possibilities,

chromosomes

using

chromosomes

were

DNA

first

in rttlOl A cells suggests that

elongation

we

was

DNA fibers

significantly

To discriminate

replication

combing (Herrick and Bensimon 1999).

prepared

from

linearly arranged

antibodies to visualize

synthesis.

have monitored the

wild-type

after release from a-factor into 0.2M HU

2003). The DNA

of DNA

replicated

(Figure 2B). Interestingly, reduced in cells

on a

DNA and

and rttlOl A cells

described

as

glass

anti-guanosine

the number of

lacking RttlOlp,

of individual For this purpose,

pulsed

previously (Versini

slide and

is either

RttlOlp

probed

with BrdU et

al,

with BrdU

antibodies to counterstain

firing replication origins

was

and the average distance between

40

increased to 68.8 Kb

replicating origins 2C).

In order to

measure

cells

exponentially-growing tracks

the

was

analyzed by of

length

replicated

which is reminiscent of

RttlOlp

is either

Although

the

the rate of were

DNA

pulse-labeled

sgslA

(Versini

mutants

indirectly

cells have

54.2 Kb in

~50%

longer al.

et

needed for efficient

increased

replication

Rtt101p is synthetically-lethal with but they do not share overlapping

proteinaceous

replication pausing

areas

throughout

which often leads to unstable

restart

mutants,

as

the

a

at

combined with rrm3A

(Fig. 3A),

determine if the cullin,

RttlOlp,

region, CEN4, found that

zones, we

that

pausing

however there

fork

speed,

pausing

intermediates that

mutants, DSB

can

repair

was

also involved in

2D

synthetic-lethality

mutants,

also

et al.

preventing

require Rrm3p

was

increased in the rrm3A deletion strain

difference in the amount of

pausing

or

replication

repaired

synthetically

shown

to

,

be converted into

to

and

lethal when

2004). To

extensive

gel electrophoresis analysis

previously

in

likely

is enhanced in rrm3A cells

previously reported (Tong

was

performed

(open arrows).

most

intermediates need to be stabilized,

as

program is

replication origins.

was

was no

and rttlOl A cells

replication

many non-nucleosomal

restarted, respectively. We confirmed that rttlOlA

proteinaceous

DNA-replication.

result rrm3A mutants confer

checkpoint

replication

We conclude that

functions

occur

replication

(Fig, 2D),

replication helicase Rrm3p

the

the genome. The

toxic double stranded breaks. As when combined with either

sites

of BrdU

Surprisingly,

in rttlOl A cells

2003).

order to compensate for the decreased number of active

Natural

length

with BrdU and the

is weak in rttlOl A cells, the an

wild-type cells (Fig.

in wild type and rttlOl A cells. was

or

to

the level of individual molecules,

at

DNA stretches

directly

as

elongation

combing

replication delay

clearly perturbed

compared

on a

pausing

at

centromeric

prevent pausing. We

(Fig. 3B,

black

arrow),

between wild type cells

41

B

H

ÏSftSs'

W0:

fl> :.ï:iii;,ï5 IIS SUSI: §ij#ïS $;$$$

Sffilsi ffl: @;llï

%-

i

;

fi '

P-value

sequence

observed

IGACTCAT[1]

25/32

1 67E-06

ATGAGTCAfJI

24/32

2.45E-06

Gcn4p-myc

TGACTCGfH

27/32

0.009

a-actin

TGACTCA

13/32

0 01

120 —

ut

c

J8 £

wt pORF-lacZ

QuriA

MORF4-onfy-lacZ

100 ,uv

t

I

80

gcn4 A nun'A/gcn4A

60 40

20 0

100 riM

rapamycin

min

0

4p-myc a-actin

30 >



aa-starvation

uriA

wt

60

«*

—-

0

30

»>#| —|

uriA

wt

60 It'

-

hrs

Gcn4p-myc elF2a

2

0

0

2



'

»f^>#-

scUrip represses GC/V4-dependent transcription. (A) Conserved motifs within the consensus site for Gcn4p is underlined. mismatches. P-values correspond of allowed Numbers in parenthesis refer to the number Fig.

4.

promoter region of scUrip-dependent genes. The

to

the false positive probabilities for each motif. (B) Extracts from indicated strains were analyzed by Western blotting using anti-myc antibody 9E10. (C) Indicated strains harbouring or the mutant, GCN4uORF4only-lacZ, were analyzed for lacZ activity. (D) Standardized mRNA expression values of GAP1, GCN4, GLN3 and GAT1 in the indicated strains relative to wild-type were obtained from genome wide transcription analysis. (E) Wild-

GCN4uORF-lacZ

type and uriA mutant cells were analyzed for Gcn4p-myc expression following 100 nm rapamycin treatment. (F) Wild-type and uriA mutant cells were analyzed for Gcn4p-myc expression following

amino

acid starvation.

96

B

PPase

+

Hrs: 0 -URI-P

+

1

0

8

:ÉiL-ÉfflMx ;jmWMm mP ^P „

"URI

+

insulin

Jtamtii,

insulin,

rapamycin 12

4

8 -

tfwwm

^f WE

ÜI

^

*ääm

URI-P

-URI

D +

vector

+insulin. rapamycin +

-

HA-mTOR(WT) HA-mTOR (RR)

-

-

-

-

-

-

+

-

•ÊltËj^À

+

-

àaut .aaatfiaE,. _o—,1

-URI-P

-URI

i«w»

URI HA-mTOR

-

0

30

1.5

60 1.0

0

30 60 120

0.5

.I

Ctr

120

m "»

o -o x o"0 CD 30 W I- 33

-H » W O

BS OD

i»n

^ i o r- tj s*

Z X ^i

siURI

w

S-

M b

ut

3'

O

Fig.5. URI participates in mTOR signaling. (A) HEK 293 cells were serum-starved, treated insulin, processed for lambda phosphatase treatment and analyzed by Western blotting for URI. (B) Serum-starved HEK293 cells were induced for the indicated time with insulin in the absence or presence of rapamycin and analyzed by Western blotting for URL (C) Untransfected HEK293 cells or cells transfected with HA-mTOR(wt) or rapamycin resistant mutant HA-mTOR(RR) plasmids were serum-starved, treated with insulin and rapamycin and with

analyzed by immunoblotting. (D) pSuper

or

pSuper-si-URI

profile of pSuper-HeLa Silencing

were

analyzed by

cells and

of URI affected the

Cell extracts of HeLa cells

pSuper-si-URI

rapamycin

stably transfected with either (E) Gene expression

immunoblotting for URI. cell lines

following rapamycin

treatment.

response of 28 genes shown in the cluster

analysis.

97

References

Sonenberg, Proc.

1.

Raught, A. C. Gingras, 98,7037(2001).

2.

T. Schmelzle, M. N. Hall, Cell 103, 253

3.

T.

4.

J. S. Hardwick, F. G. Kuruvilla, J. K.

B.

N.

Natl. Acad. Sei. USA

(2000).

Peng, T. R. Golub, D. M. Sabatini, Mol. Cell. Biol. 22, 5575 (2002).

Shamji, S. USA96,14866 (1999).

Schreiber, Proc. Natl. Acad. Sei.

Tong,

A. F.

5.

J. Lisztwan et al., EMBO J.

6.

H.

7.

Siegert, M. 621 (2000).

8.

I.E.

9.

K.

10.

J. Martin-Benito et al., EMBO J. 21, 6377

11.

Woychik, 313(1990).

12.

M. Kanemaki et al., Biochem.

13.

M. A.

14.

D.

15.

Y. Murakami et al., Nat. Genet.

16.

T. Ito et al., Proc. Natl. Acad. Sei.

USA9S, 4569 (2001 ).

17.

C. J. Gimeno, P. O.

A.

Sutterlüty

Vainberg

Siegers

17, 368 (1998).

al., Nature Cell Biology 1, 207 (1999).

et

R.

L.

R. Leroux, C. Scheufler, F. U. Hartl, I. Moarefi, Cell 103,

et al., Cell

93, 863 (1998).

et al., EMBO J. 18, 75

S. M. Liao, P. A.

N. A.

(1999). (2002).

Kolodziej,

Biophys

R. A.

Young,

Genes Dev. 4,

Res. Commun. 235, 64

(1997).

Wood, S. B. McMahon, M. D. Cole, Mol. Cell 5, 321 (2000).

Dorjsuren

et

al., Mol. Cell. Biol. 18, 7546 (1998).

10, 261 (1995).

Ljungdahl, C.

Styles,

G. R. Fink, Cell

68,1077

(1992). Krön, N. A. Gow, Curr. Opin. Cell. Biol. 7, 845 (1995).

18.

S. J.

19.

A. P. Gasch et

20.

K.

21.

V. A.

Natarajan

al., Mol. Biol. Cell 11, 4241 (2000).

et al., Mol. Cell. Biol. 21, 4347

Cherkasova,

A. G.

(2001 ).

Hinnebusch, Genes Dev. 17, 859 (2003).

98

22.

J.

Chen,

X. F.

Proc. Natl. Acad.

Zheng, E. J. Brown, S. L. Schreiber,

Sei. US4 92, 4947

(1995).

Acknowledgments:

23.

laboratory for discussions and Drs. A. G. Hinnebusch for myc-GCN4 and lacZ-reporter plasmids, G. Thomas for mTOR plasmids and R. Agami for the pSUPER vector. Special thanks goes to Drs. J. Paszkowski, U. Müller and G. Thomas for critical reading of the manuscript. M. V. acknowledges support We thank all members of

our

from the Association pour la Recherche sur le Cancer. This work is supported by the Swiss Cancer League, the Novartis Research Foundation and an SNF grant to M. P.

SUPPORTING ONLINE MATERIAL Materials and Methods

Yeast

two-hybrid

screen

carried out with the

yeast reporter

(pGAD-GH plasmid; Clontech) 113-435 of human SKP2 clones

were

sequence construct STAP1 in STAP1

analyzed by

encoding

cloned into

digested

A standard

Y153, using

two-hybrid

a

Gal4-fused SKP2

a

direct DNA

sequencing.

as a

bait. Positive

an

was

untagged

used to version of

generate pcDNA3-HA-STAP1, pcDNA3-

with BamHI and EcoRl and the

HA-tag previously

an

acids

All were found to contain a

expression plasmid encoding

pcDNA3 containing

library

encoding amino

ORF of 157 amino acids. One

To

screen was

HeLa cDNA

expressed from pAS2-1 plasmid

pcDNA3 (InVitrogen).

was

strain

and

predicted

a

mammalian

a

plasmids.

and

resulting fragment introduced at the

Hindlll/BamHI site.

URI to

peptide sequences obtained from MS/MS sequencing,

identify overlapping

EST sequences and human

encoding full-length URI. Based and used in PCR reactions to

from

a

human

amplify

the first 426

lymphoid library (Strubin

product was digested

et al.

used

genomic sequences

this information,

on

were

primers

were

designed,

bp of URI coding sequence

1995).

The

resulting

with BamHI/EcoRI and subcloned into

PCR

pcDNA3.

The

commercially available cDNA clone IMAGE clone 3609351, containing the 3' part of the URI cDNA,

was

digested

with EcoRl and the

resulting fragment

99

into the EcoRl site of

ligated

region of plasmid

pcDNA3 containing

resulting plasmid is referred

URI. The

Flag-epitope tagged

a

version

was

426bp fragemnt of the 5'

the

to

pcDNA3-URI.

as

from

HeLa cDNA

a

cDNAs

encoding

amplified

from

human cells

yeast proteins scPfd6, scRpb5

the

cDNA libraries

pGEX

we

(Brummelkamp

were

PCR

amplified

library (Clontech) and subcloned into pGEX and pcDNA3.

yeast

subcloned into

in the

generated, resulting

plasmid pcDNA3-FT-URI. Human RPB5 and p17 cDNAs

From this

and

a

For stable

following oligonucleotides

2002)

to obtain

scUrip

were

one-step PCR protocol and

pBS-KS, respectively.

cloned the et al.

using

and

into

silencing of URI in

pSuper

pSuper-siURI:

forward:

GATCCCCATGGAGAAGATACGACATCTTCAAGAGAGATGTCGTATCTTCT

CCATTTTTTGGAAA; reverse:

AGCTTTTCCAAAAAATGGAGAAGATACGACATCTCTCTTGAAGATGTCGT ATCTTCTCCATGGG.

Antibodies. Mouse monoclonal antibodies 4H7 and 8WG16 directed

respectively,

RPB5 and RPB1, al. 1989;

Nguyen

and CUL1

were

et al.

(105.72

and

105.128)

previously (Lisztwan

et al.

and URI

(179.30

and

179.63)

human

proteins expressed

(GST)-fusion protein.

Rabbit

polyclonal

a

as

antibodies

were

Mouse

raised

glutathione

recognizing

against

the

S-transferase STAP1

against URI(179.63.1,179.58.1)

against full-length human proteins expressed

Polyclonal

1998).

were

maltose-binding protein (MBP)-fusion protein.

Monoclonal antibodies directed raised

et

recognizing SKP1 (90.122b), SKP2 (95.60.2), STAP1

respective full-length

generated against

previously (Thompson

have been described

1996). Polyclonal antibodies directed against SKP2

described

monoclonal antibodies

against

antibodies

recognizing

URI

were

as

have been

GST-fusion

proteins.

provided by Dr. Seishi Murakami

100

(Dorjsuren were

et al.

raised

1998).

Rabbit

polyclonal

anti-TIP48 and TIP49-antibodies

TIP49

(TIP48 peptide: MKEYQDAFLFNELKGETMDTS;

LFYDAKSSAKILADQQDKYMK). Peptides hemocyanin by glutaraldehyde coupling polyclonal number:

antibodies

Q9NUG6)

specific

were

peptide:

into rabbits. Rabbit

injected

and

incubation with first

a

against bacterially produced

GST-affinity-column

followed

protein column, prepared by covalently cross-linking

the

a

achieved

sera was

GST-fusion

1999).

described

(Lisztwan

et al.

Anti-peptide

were

affinity-purified

as

Monoclonal antibodies HA-11, 12CA5, 9E10 and anti-actin mAb

purchased

form Babco,

International

Boehringer Mannheim, Santa Cruz

respectively. Anti-elF2a antibodies

were

to

respective proteins and Lane

with

1998).

by

GST-fusion

dimethylpimelimidate (Harlow

glutathione-Sepharose antibodies

Accession

p17 (PFD4r,

protein. Affinity-purification of the different polyclonal rabbit

by

C-termini

coupled to keyhole limpet

were

for human PFD3 and

raised

respective

to their

against peptides corresponding

were

and Chemicon

kindly provided by

Dr.

Ron Wek.

Immunoprecipiation, immunoblotting, gel filtration, phosphatase assay. Immunoprecipitations and immunoblotting described

(Lisztwan

et al.

membranes has been

lysis

buffer

Ponceau S

as

of western blot

procedures of the

109

HeLa cells

[50 mM Tris-HCI (pH 8.0), 150

mM

NaCI, 1% Triton X-100,1

Superose

for 30 min

6

gel-filtration

on

ice, lysates

min at 4°C at 3000 g and filtered

Superose

(Sigma) staining

to the

performed according

3-4

mM DTT, 0.5 mM PMSF, 20 mM NaF and 1

(Roche)]

performed essentially

x

manufacturer. For 10 ml

1998).

were

6 column

0.5 ml/min flow rate

were

through

(FPLC; Pharmacia). using gel-filtration

x

were

lysed

in

protease inhibitor cocktail

pre-cleared by centrifugation for 30 a

0.45

urn

filter

Gel filtration

buffer

[50

mM

prior

was

to

loading

performed

on a

at 4°C at

Hepes (pH 7.5), 150 mM

NaCI, 10 mM MgCfe and 0.5 mM DTT] and 4 ml fractions

were

collected. Blue

101

Dextran 2000

(232 kDa) For

(158 kDa)

and aldolase

phosphatase experiments

DMEM w/o

buffer %

(2 MDa), thyroglobuline (669 kDa), ferritin (440 kDa),

[50

NP40]

serum

used

as

for 30 min

(NEB)

0.1 mM EDTA, 5 mM

(pH 7.5),

on

ice. 200 ug of

in 50

pi

Reactions have been

standards.

HEK293 cells have been starved for 24h in

prior

to

lysis

DTT, 250 mM NaCI. 0.5

1x X-PPase buffer

[50 mM Tris-Hcl (pH 7.5), 0.1 for 30 min at 30° C.

Brij 35]

stopped by adding 20 ul Laemmli-buffer and boiling for

loading

6% SDS-PAGE

on

gels.

Purification of STAP1-associated proteins and

purification

of STAP1-associated

mass

proteins,

spectrometric 4 ml of

packed

analysis.

For

cells

were

lysed for 30 min

7.5),

250 mM NaCI, 5 mM EDTA, 0.5% NP-40, 50 mM NaF, 0.2 mM

(DTT),

1 mM dithiothreitol

ug/ml aprotinin], lysates

cleared

in PL

protein extracts have incubated with 400

mM EDTA, 5 mM DTT, 250 mM NaCI, 0.01 %

10 min before

molecular-weight

and induced with 10uM insulin for 1h

mM Tris-Hcl

U of X-PPase

were

catalase

dounce

on

ice in 50 ml TNN-buffer

1 mM

by centrifugation

fluoride

phenylmethylsulfonyl

homogenized (10 at 18000

strokes with x

mM Tris-HCI

[50

a

B

g for 20 min.

(pH

Na3V04,

(PMSF),

pestle)

HeLa

10

and

Supernatants

were

collected and incubated for 2 hrs at 4°C with anti-STAP1 mAb 105.128

covalently coupled to protein A-Sepharose using

coupling procedure. Immunobeads proteins eluted with 300 K2HPO4, concentrated a

6%-15%

protein

ul 0.2 M

on a

were

washed 4

were

analysis. Peptides

dimethylpimelimidate x

with TNN-buffer,

glycine (pH 2.5), neutralized with 100 ul 1M

Centricon concentrator

gradient SDS-polyacrylamide gel.

bands

a

The

(Amicon)

gel

was

and resolved

on

silver stained,

excised, digested with trypsin and subjected to sequence

were

sequenced using nano-electrospray tandem

spectrometry (NanoESI-MSMS)

on a

API 300

Toronto, Ontario, Canada) and identified Identities of STAP1-associated

proteins

as

mass

spectrometer (PE Sciex,

described

were

mass

(Mann

confirmed in

an

and Wilm

1994).

independent

102

experiment using

MALDI-TOF

analysis

on a

TofSpec

2E

(Micromass,

Manchester, UK) according the direction of the manufacturer. The full dataset of this

analysis

In vitro

is available upon request.

binding

expressed

reactions.

GST-fusion

as

STAP1, hRPB5, scRpb5

proteins

in E.

coli, purified

Sepharose beads and subsequently incubated vitro translation

(Promega) analyzed

as

on

products

previously

described

on

scPfd6

glutathione-

35S-methionine-labeled

(Krek

et al.

1994).

Bound

in

proteins

were

by fluorography.

have been used

techniques. The following yeast strains

study: yBM80 (W303):

were

scUrip generated by the TNT system

or

SDS-PAGE and detected

Yeast strains and in this

of either URI

with

or

uri::HIS3 ade2-1

trp1-1

can1-100

Ieu2-3,112

his3-11,15 ura3 psi+. yBM79 (W303):ade2-1 trp1-1 canMOO Ieu2-3,112 his311,15 ura3 psi+. yBL45 (K699): ade2-1, trp1-1, can1-100, leu2-3,112, his311,15, ura3, psi+, ssd1-d2 bar- uri::Uri-13myc His3. yBL284 (W303): ade2-1,

trp1-1, can1-100, leu2-3, 112, his3-11,15, ura3, psi+ gcn4::kan. yBL285

(W303): ade2-1, trp1-1, canMOO, leu2-3, 112, his3-11,15, ura3, psi+, gcn4::kan uri::HIS3.

YBL295

(W303): ade2-1, trp1-1, can1-100, leu2-3,112,

his3-11,15, ura3, psi+, ssd1-d2, uri::HIS gcn2::kan. yBL282 (W303): ade2-1,

trp1-1, can1-100, leu2-3,112, his3-11, 15, ura3, psi+, uri::Uri-13myc-HIS3. yBL225 (2):ura3. yBL234 (2):ura3-52

uri::KAN.

yMJ205 (W303): cdc34-2

ade2-1, trp1-1, can1-100, leu2-3,112, his3-11, 15, ura3, psi+, ssd1-d2. Details about strain constructions will be otherwise standard used

(Guthrie

at 25 °C until

yeast growth

and Fink

1991b).

early-log phase

in

provided upon request. If

conditions and

For

protein

not stated

genetic manipulations

half-life

analyis

synthetic media. Cells

cells

were

were

were

grown

then shifted to

a

non-permissive temperature of 37 °C for 3 hours. At time 0 cyclohexamide was

added to

a

final concentration of 50

ug/ml

and

samples

were

taken at the

103

extracts

indicated times,

protein

intensities

analysed

were

with

were

analyzed by

Quantity

One

was

as a

ratio

(intensity

always normalized and

set to

For

a

where time 0

value of 100. Standard deviations

experiments

and

plotted

on a

scatter

performed

graph.

Hinnebusch).

(Mueller

under the

tap

1986),

and Hinnebusch

Nasmyth 1985). Yeast cells

were

water after

growth

on

were

(OD600

=

washed twice with H2O

in either amino acid starvation medium or

described

for two

in

urea

medium

(Breeden

growth by washing

days

at 30°C. In nutrient

cells grown in YPD

diluted to OD600 of 0.1 and grown in YPD to

0.4),

glucose),

were

yeast

experiments, overnight cultures of yeast

medium

as

tested for invasive

extract/peptone/dextrose (YPD) plates starvation

LacZ assays

where the ATGs of uORF1-3

GCN4uORF4only-lacZ,

or

have been mutated

yeast plates

Alan

yeast strains transformed with the reporter plasmids

on

GCN4uORF-lacZ

prewarmed

early log-phase

at 30°C and

resuspended

(3.4 g/l yeast nitrogen base,

(3.4 g/l yeast nitrogen base,

3%

5 uM urea, 3%

glucose, appropriate supplements for auxotrophs Ura, Leu, Trp, Ade). diauxic shift

0 hrs time total

experiments yeast

point)

protein

(yBM79) 30°C

and

were

were

grown to

early log phase

taken at different time

in YPD

(=

points thereafter for

expression analysis. Three colonies of wild-type strain

and uriA

on a

aliqots

cells

In

extraction.

Global gene

to

were

expression of GCN4p-myc yeast cells have been transformed with

plasmid p3330 (kindly provided by

and

version 4.2

determined and

were

GCN4-myc/control intensity),

of

determined from three individual

blotting. Band

quantitation software,

background

from Biorad. Intensities minus the lane

expressed

Western

(yBM80)

were

inocculated in 10 ml YPD, grown

overnight

at

shaker at 225 rpm, diluted to OD600 of 0.1 and continued to grow

early log phase (OD600=0.3).

Yeast cells

were

for 5 min at 3000 rpm and total RNA extracted

by

collected the

by centrifugation

acid-phenol method

104

from wt

(yBM79)

(yBM80)

In order to define

1990).

al.

and uriA

RNA from six

rapamycin

time-course

(ODeoo YPD.

of

0.3)

Aliquots

rapamycin

and

of 50 ml cell

points by centrifugation frozen in

yBM79

0.3),

and

collected

were

medium at

a

yBM80 yeast

cells

by centrifugation

30°C, (0.34

described above.

were

grown to

pellets

were

snap-

using

glucose). Aliqots

following oligonucleotide

extracted

to the was

were

performed

on

protocol

Ten ug of total RNA

recommended

transcript labelling

kit

protocol. The cRNA

cleaned

phase

In vitro

the Enzo

centrifugation

x

recommended

one

by Qiagen.

minute

by was

BioArray High Yield RNA the manufacturer's

using RNAeasy clean-up

g and then left for

removed

transcription

columns

improve the recovery from the columns, the elution water

the matrix at 27

by Affymetrix.

Double-stranded cDNA

(Enzo Diagnostics, USA) following

was

was

priming: 5'-

(Eppendorf).

template using

as

system for cDNA synthesis

and the aqueous

Phase-lock Gel

1 ug of cDNA

suspension

performed using YG_S98

Choice

used for

by phenol:chloroform

centrifugation through

of 50 ml cell

points for extraction of total RNA

SuperScript

the

temperature and washed

resuspending them in amino acid starvation

Microarray analyses

transcribed

early log phase (OD600 of

ggccagtgaattgtaatacgactcactatagggaggcgg-(t)24-3'.

To

early log-phase

independent replicate cultures

at 3000 rpm at room

% YNB, 3%

(Life Technologies) according

was

were

harvested at indicated time

were

GeneChips (Affymetrix, Santa Clara, USA).

The

URI,

final concentration of 100 nM in

amino acid starvation two

harvested at indicated time

reverse

loss of

yBM80

and

grown to

for 5 minutes at 3000 rpm and

twice with H2O at 30°C before

were

yBM79

by

experiments (Fig. 5B), two separate

added to

suspension

affected

et

liquid nitrogen before RNA isolation. For measuring global gene

expression following of

was

of

cells

replicate cultures of yBM79 and yBM80

microarray analysis (Schmitt

expression

gene

independent replicate cultures

For the

analyzed.

global

strains for

was

(Qiagen).

spun into

prior to the standard 8000 g

This low

speed wetting step gave

105

nearly

double the

cRNA

was

fragmented by heating

(pH 8.1),

acetate

of eluted RNA

yield

Affymetrix.

fragmented

Ten ug of

staining

protocol The

was

performed

standard in

were

Suite v4 for

AvgDiff

percentile

an

perform

The

values remained

negative

expression

calculations

were

or

based

on

AvgDiff values

whose absolute fold

the lower 90% confidence bound for fold

equal

provided

(Silicon Genetics)

used to

was

changes

Fig.

of human cells

respective

onto

a

10cm dish 24h

RNA was

prepared from

the TRIZOL the

prior

on

essentially

as

HG-U133

v5 and

assigned

Analysis

GeneSpring

5.0

of the

to functional

GeneSpring

4.2

for the presence of

global transcription analysis

HeLa cell line

samples

chips

from

was

were

seeded

rapamycin. Total

were

oligonucleotide microarray chip

described above.

MicroArraySuite

were

two 10cm dishes at the indicated time

RNA

and

and had P values for

change

4A. For

GeneChips

greater than 50, that

treatment of the cells with 20 nM

reagent (Life Technologies).

analyis

to account for

analyze promoter regions

million cells of the

50th

all arrays. Fold

at http://casiama.c8hi.ora/iian.

conserved sequence motifs shown in one

were

to 0.05. Yeast ORFs

groups using the tools

calculate fold

for those genes

between individual

passed

or

across

signals

only genes

smaller

(AvgDiff)

very low have been set to the

normalized

intensity

scanner.

are

average difference

include

change

(Affymetrix) using the

performed using the

values called absent

variations in overall fluorescence

fold

hours). Washing

16

Affymetrix GeneChip

determining

analysis.

mM Tris-

YG_S98

a

experiments performed

cluster

of the

hybridised to

was

The

recommended by

(www.biostat.harvard.edu/complab/dchip/) to

values and dCHIP v1.0 and to

as

procedures (45°C,

request. Chip analysis

Affymetrix Microarray

changes

MgOAc]

Fluidics Station 400

a

data sets for all array

communicated upon

change

cRNA

EukGE-WS2v4 and scanned in

complete

whose

fragmentation buffer [40

in 1x

100 mM KOAc, 30 mM

GeneChip (Affymetrix) using and

(E. 0., unpublished observation).

points using

processed for Affymetrix

performed using

(Silicon Genetics). Changes

in gene

106

expression

were

determined

that pass

replicates

signed

a

by looking

Wilcoxon rank test. The

was

identify

list of 194 genes that treatment in

rapamycin

Welch t-Test

(pwb[

2

3 4

WBf

I

^HA-STAP1 12

1

P

3

4

—>^HA-STAP1

Fig. S1. STAP1 is a member of the a class PFD family and part of a multiprotein complex. (A) Amino acid sequence alignment of STAP1 with members of the prefoldin protein family. Secondary structure elements were deduced from the crystal structure of M. thermoautotrophicum PFDcc subunit (Siegert

2000) and are represented as blue cylinders (a helices) and grey arrows (P strands). (B) Characterization of anti-STAP1 antibodies. Whole cell extracts of U2-OS cells (lanes 1 and 3) or U2-OS cells expressing untagged STAP1 (lanes 2 and 4) were analyzed by Western blotting with anti-STAP1 mAb 105-72 or affinity-purified polyclonal anti-STAP1 antibody. (C) Nuclear extracts of 293 cells were subjected to immunoprecipitation with control antibody (mouse IgG, lane 1) or anti-STAP1 mAb 105-128 (lane 2) coupled to protein A-Sepharose and the immunoblot probed with anti-SKP2 or antiSTAP1. (D) Top: Extracts from untransfected U2-OS cells (lanes 1 and 2) or from U2-OS cells transfected with an HA-STAP1 expression plasmid (lanes 3 and 4) were immunoprecipitated with control antibody (preimmune serum, lanes 1 and 3) or polyclonal anti-SKP2 antibody (lanes 2 and 4) followed by Western blotting using anti-HA antibody. Bottom: Extracts were immunoblotted with anti-HA antibody before immunoprecipitation. et al.

108

2MDa

Fraction: 1

5

440 kDa

670 kDa 15

10

20

158 kDa

25

30

35 STAP1

-!

-

PFD3

mi

URI RPB5

TIP49

TIP48

SKP2 CUL1

SKP1

t3

2MDa

2

440 kDa

670 kDa

6

14

18

22

26

158 kDa 30

34 STAP1 URI RPB5

TIP49 TIP48 STAP1 IP

Fig.

S2. STAP1 and URI

whole cell

lysate

are

part of

of HeLa cells

was

processed

for

~

WB

1MDa

complex in human cells. (A) A on a Superose 6 gel-filtration Western blotting using

fractionated

processed for proteins. (B) Fractions shown

column and individual fractions antibodies to the indicated

a

>

with anti-STAP1

immunoprecipitation immunoblotting with antibodies directed against right.

the

(A) were antibody followed by proteins indicated on in

the

109

complex by co-immunoprecipitation. (A) Aliquots immunoprecipitated using anti-URI polyclonal anti-TIP48 (lanes 2), polyclonal (lane 4) coupled to proteinA sepharose or control antibody (lanes 1 and 3) and analyzed by immunoblotting with antibodies directed against indicated proteins. (B) Prefoldin subunits STAP1 and PFD4r are stoichiometrically associated with URI. Whole cell extracts prepared from HEK293 cells were subjected to immunoprecipitation using monoclonal antibodies directed against human URL Lysates before (lane 1) and after sequential immunoprecipitations (lanes 2 and 4) and immunoprecipitates (lanes 3 and 5) were analyzed by Western blotting using antibodies to the proteins indicated.

Fig.

S3.

Analysis

of the URI

of HEK293 cell extracts

were

110

UH (hü) UR! St) UH Wra) URI

(EC)

URli&Cj Pf 05 (»C)

Ji*«ArTV*»rPDrsFP^JLp-JLPAfcvPlLft*riWlW]UM!UEI^ 1. HEP.PAK«W*ve»FKR *;*nHXHVAic#kxiWii().QKgK.:Bi.F.. HKABrVKKD. .BE VmKt*S*iï»tKH*ï>flI.tJl'I»iTnEPfii,>iftXi*«aff*'J.">lt(¥«i:aBTF.Bi.l'ÔlS«Se lia-.axKWtrriiBttuV. Htt».HICBAPlPB>F]|IVBI.R DLpBLOK ScrïltïVBttïOï:, ^HEKLCÏUBRKKEKKB.gNPtCNRFTJlCltpELHlinpDIIiriKfWBUTJt*^*!.^» IB lia.PPltHîf'AOPBPStVB URllcfi) RU»ïBVFfrB^"*V&**IiV*tïl>AULVlMlKI«I AHHOEDFLPBK«**ï»LPP, ,«DOU»lJ&OHÏUft»BflD«»Hll^LOVCI>HSiP7!3ÏFSHÏVEPKRVflIWTBnF

123

i

22-1

:

JlB

.234 14 L

.,,,,

'3H4

.»»LAr^KKfiEAKRÄÄRKßTCSDHSAftE^PTI

-«5

houvkvckok^i

(dm) iS4.»XWKFI.CAtHftlitl£AÖiltl... .öaCABCBLKT»BKLVRttliMiOETEtP6fl

KBBTA3rWRNVBI8»ri3BS*lafiJ>ll»lJS*bBi6fn>KH

,OK«MIFfl««FAI.I,AOFlï¥TOKic>IF.m.LD "fel»bDlllïïft*iltïn*I»».*lti.iBrKV,Ml»ÏÉE^^ I.r,«VEEVFOBPT

URlfce)

2IB

URIfflC Pf 05 (se)

"5

URIfht.) URI(flt)

jAïLKTPADZXnArv»tfVHUïTK)-iiKKiLKSKaHjlliWttVCS»£BE,«SAk«F|ipa(i(;uLTis^Rrici:ATCSDTViEsil1lfx.

URl(rfm)

Jsû:irvvvMÏ.»[ji)LBiHi.'>*fœBlBl,BAS

"ft,

1

-

URI (dm)

Uffl

13«

fliJlSÄHAKttPl. 1J7 9*>*«t*vHDi.iriH>» ftltlMK. J.PS MWt

KlhïAFIÏAKICBVOtr^^HirYAÏffflVTKEH^ftïfffX^ïlI-tlVUnibfiR«

.Eft. .AKEÏBlt

.

t

.

FLRAOfcBVLF.AF.HBKIOKVBPiaööLiaLM

UtfJlJ^fiEOHÄLLi'HEÜSEP -

i

(



TH*Ä«H3t>tta[.«fl«'*'T*T9VHtVF!!STV»:F.Tim

*V»TI.l>SP[lïASni)W ,

i;;;::,,:::;;.-^

URI

IfiO)

URI (SC

epgjlt*K|ili{Ll,FLitïiPllArKiiT#ïit*BÎÏ'Vflril,TPJPJiiJf*iiP»LPTiPi:Km*l/i.iJ

265

,

EKBKltVVÎtfWX-L.'i.

334iVTKTTVKRKHS&VRFA«ptEH

miOESTflÎANKMTHDIFDÊDt-FAK

NLU

«PSKBI-flTVHC. ekthkbrtj,

vK*F»Ot((T-v*lK*«nBVÏE U0»* LECÇ

VOEAOBVBBKÏ KU BLI

.

ÏRBILItatEr

DDiiKt,*. ÏPVORSAH« fai

.

,

r

AKUJïïVkËMLïlilAkKtr.

PEBHUftftMSTUTrPBf,

USAS WttBTB VÇB F" KPB ASF.FAKTl.BTr

.

SKÜliALi(APl,)£I-AflD^PAii «TU IVRF)(Pï(Ç)CEPa* = AOl>At>VITfi

IpTU Vt.BI VVBKÄP VVODlJl AÏ» V VÜÜ V I EKB

PVUOPV

PFD5(BCJ -

1

URI

(dm)

URI ICC) UWlw

.airgOlBfil» ,

I KXEPAVTD

,

.

ïVcAI

-

:PEOEEHVKVPIrtaEAAXliaAÏBPPPaor3KriiS9EiCr¥a!ia.

.LLHRIHGEaTQLIrt..TLHI3FTHS..ffAIP5PSBp(}3P

ivsipsvicfu,

1

-

-

...

.

SIIPTâKdOSPPT(

4331

SlS-rVvnoiîE

ÉP'ïinÙrVBKBHtlItSKaefB'Ii-FKKKKLKSLCKPKSSKS«

KKFPP«T3,ENT3n DnTODnnDOKKitLL«HK

KHH2DltßUHPP3l{ltt£V