GENOMIC INSTABILITY AND CANCER: INSIGHTS FROM ANALYSIS OF THE

BLOOM’S SYNDROME HELICASE

Ian D. Hickson

CR-UK Oxford Cancer Centre,Weatherall Institute of Molecular Medicine,

University of Oxford.

PRESENTATION OUTLINE

• How do cancers arise - what genetic changes are necessary?

• Genomic instability and its role in tumorigenesis

• Bloom’s syndrome as a model for understanding how aberrant mitotic recombination can drive tumorigenesis

• A conserved pathway involving the BLM protein and its partners topoisomerase IIIα and BLAP75/RMI1 for resolution of homologous recombination intermediates

• Cancer is a genetic disease - but it is polygenic

• Results from defects in genes that control cell birth and/or cell death

• Cancers acquire several capabilities in order to be life-threatening- evade apoptosis- insensitive to anti-proliferative signals- ability to invade tissue and metastasize- immortal (no replicative limit)

HOW DO CANCERS ARISE?

1st mutation 2nd mutation 3rd mutation nth mutation

CANCER

THE CLASSIFICATION OF CANCER GENES

• ONCOGENES•Mutation renders gene constitutively active (e.g. Ras)•Drive tumorigenesis e.g. by making cells independent of mitogenic growth signals•One activated allele sufficient

• TUMOUR SUPPRESSOR GENES•Mutation renders gene inactive (e.g. Rb)•Loss of function permits unregulated cell cycle progression etc.•Usually both alleles inactivated

• GENOME STABILITY GENES (CARETAKERS)•Mutation renders gene inactive•Not necessarily directly involved in or rate limiting for neoplastic transformation•Not selective; simply increase probability that oncogenes/TS genes will be hit•Usually both alleles inactivated•Particularly potent when defect inherited

Oncogene and TS gene changes directly drive neoplastic transformation by permitting cell proliferation and/or abrogating cell death (GATEKEEPERS)

Maintenance of Genomic Integrity

Ataxia telangiectasia(ATM)

Fanconi’s anaemia(FANCA-M)

Ataxia telangiectasia-like disorder

(MRE11)

Nijmegen breakage syndrome

(NBS1)

Hereditary breast and ovarian cancer

(BRCA1/2)

Xeroderma pigmentosum

(XPA-G)

Bloom’s syndrome(BLM)

1st mutation 2nd mutation 3rd mutation nth mutation

CANCER

normal cells malignant cells

Bloom’s syndrome(BLM)

Bloom’s syndrome

- Autosomal recessive disorder

- Short stature

- Skin abnormalities

- Male infertility/female subfertility

- Predisposition to cancer

RARE TUMOURS

Age at Diagnosis

Cases

Medulloblastoma1

Metastatic to liver1

Lung1

LYMPHOMA

LEUKAEMIA

CARCINOMASkin11

0 10 20 30 40 50

2 Wilms's tumor

5 Uterus

13 Large intestine

7 ALL3 Osteogenic sarcoma

5 Pharynx3 Oesophagus

7 Breast

3 Larynx2 Stomach

2 Hodgkin's disease20 Non-Hodgkin's14 ANLL

The first 100 Cancers in the Bloom's Syndrome Registry

CANONICAL DOMAIN ORGANIZATION OF A RecQ HELICASE

Regulatory domain NLS

HELICASE RQC HRDC

Catalytic ‘core’

7 motifs of SF2 helicases

DOMAIN STRUCTURE OF SELECTED MEMBERS OF THE RecQ HELICASE FAMILY

HELICASE RQC HRDC

HELICASE RQC HRDC

HELICASE RQC HRDC

HELICASE RQC HRDC

HELICASE RQC HRDC

HELICASE

HELICASE

RQC HRDC

HELICASE RQC

Exo

NLS

E. coli RecQ

S.c Sgs1

S.p Rqh1

H.s RecQ5β

H.s RecQ4

H.s RecQ1

H.s WRN

H.s BLM

TOPO III

TOPO III

TOPO III

Werner’s syndrome (WRN)

Rothmund-Thomson syndrome(RECQ4)

RecQ HELICASE DEFICIENCY DISORDERS

Normal until pubertyPremature ageing

CataractsCancer prone

Congenital skeletal abnormalitiesSparse hair and poikiloderma

CataractsCancer prone

CANONICAL DOMAIN ORGANIZATION OF A RecQ HELICASE

Regulatory domain NLS

HELICASE RQC HRDC

Catalytic ‘core’

7 motifs of SF2 helicases

OVERALL STRUCTURE OF THE CATALYTIC CORE OF E.coli RecQ

Courtesy of Dr James Keck

MUTATIONS IN BLM

C90

1Y

S186

X

K27

2X R36

4XIV

S6: 3

’SS

S595

X

Q67

2RQ

645X

Q70

0XFS

: 735

+4-X

FS: 5

14+1

-X

C10

55S

Del

E11

-12

(2)

IVS2

: 5’S

S

R89

9X

FS: 6

56+4

-X

FS: 9

74+2

3-X

FS: 1

086+

10-X

Q75

2X

G89

1E

S104

XFS

: 91+

36-X

FS: 3

30+3

-X

Q54

8XW42

8X

IVS8

: 5'S

SFS

: 750

+24-

X

W88

1X

FS: 1

074+

4-X

S109

3XR

1139

XFS

: 115

8+5-

XY1

170X

FS: 1

226+

51-X

FS: 1

195+

2-X

FS: 1

009+

23-X

C10

66Y

D10

64V

FS: 1

242+

13-X

IVS7

: 5'S

S

FS: 9

41+2

6-X

Del

E20

-22

FS: 8

02+3

-X

IVS1

8: 5

'SS

FS: 1

93+1

1-X

H96

3Y

FS: 2

56+7

-X

FS:4

48+1

-XL5

43X

W56

7X FS: 8

29+4

-XFS

: 835

+17-

X

Q90

9XG

952V

Q10

40X

C10

55G

C10

55R

Q12

83X

V119

8M

FS: 1

084+

7-X

H13

24Y

IVS9

: 5'S

SIV

S11:

5'S

S

IVS1

6: 5

'SS

Regulatory domain NLS

Del

E15

HELICASE RQC HRDC

BLM contains four distinct biochemical activities

• DNA dependent ATPase :

• A 3’-5’ DNA strand displacement activity

3’ 3’5’ 3’ 3’

5’BLM BLM

• A Holliday junction branch migration activityBLM

• A single stranded DNA annealing activity (ATP independent):

BLM

Elevated SCEs are diagnostic of Bloom’s syndrome cells

HOW ARE SCEs GENERATED ?

DOUBLE HOLLIDAY JUNCTION

Rad51-mediatedstrand invasion

Copying of genetic information from

homologous sequence

ssDNA gap

Gene ConversionEndonucleolytic cleavage and rejoining of junctions

Endonucleolytic cleavage and rejoining of junctionsin opposite orientations

Gene Conversionwith crossing over

[SCE]

‘Early’ alternate pathway(SDSA)

‘Late’ alternate pathway

TOPOISOMERASES

Class

Type 2

Type 1Type 1A

Type 1B TOP1

TOP3

TOP2

Gene (S. cerevisiae)

• Ubiquitous, highly conserved enzymes

• Act to disentangle topological problems that arise during DNA metabolism

How might BLM and topo IIIα catalyze a non-endonucleolytic resolution of double Holliday junctions?

Branch migration

Double Holliday Junction(Hemicatenane)

BLM?

Topo III?

*

Anneal oligos

R1 H1

*

*

Ligate nicks

RsaI

HhaI

1.5 helical turns

Generation of a DNA substrate containing two Holliday junctions

80-mer oligonucleotides

11bp 14bp 11bp

+ + + + + +

BLM

hTOPO IIIα

B1

BLM and hTOPO IIIα cooperate to resolve DHJ

Rsa

I

RsaI

HhaI

Requires the ATPase/helicase activity of BLM and the active site tyrosine of topo IIIα

BLM and hTOPO IIIα catalyze a novel mechanismto resolve double Holliday Junctions into non-crossover products

Double junction dissolution

BLM BLM

hTOPO IIIα-mediatedstrand passage

event

hTOPO IIIα-mediatedstrand passage

event

Non-crossoverproducts

BLM WRN BLM RECQ1 BLM RECQ5β

Double junction dissolution is catalyzed specifically by BLM

hTOPO IIIα

In contrast, any type IA topoisomerase is functional

The C-terminal domain of BLM is required fordouble junction dissolution

BLM BLM-NC

hTOPO IIIα

BLM-N

NLSHELICASE RQC HRDC

Catalytic ‘core’TOPO III

HELICASE RQC HRDC

HELICASE RQC HRDC

BLM

BLM-N

BLM-NC

C-terminal truncation of BLM disables the HRDC domain

Helicase function is not impaired in BLM-NC

NLS

HELICASE RQC HRDC

Catalytic ‘core’TOPO III

HELICASE RQC HRDC

HELICASE RQC HRDC

BLM

BLM-N

BLM-NC

The HRDC domain of the RecQ helicase familyforms an evolutionarily conserved protein-fold

Liu et al, 1999

A third subunit in the BLM/topo IIIα complex

BLAP75, an essential component of BLM-Topo IIIα complex

BLM

TOPO III

BLAP75RPA70

BLM

TOPO III

BLAP75RPA70

BLAP250BLAP250

BLM

IP

BLA

P75

IP

Yin et al. EMBO J. 2005

TOPO IIIα

BLAP75

Untr

eate

dCo

ntro

l olig

oBL

AP75

olig

o1BL

AP75

olig

o2

Beta-actin

siRNA Oligo

BLM

BLAP75 is required for the stabilityof the BLM/topo IIIα complex

Courtesy of Dr Weidong Wang

BLAP75 is conserved in yeast (Rmi1/Nce4)

I II IIISc Rmi1 241aa

I II IIISp Rmi1 235aa

I II IIIHs BLAP75 625aa

OB-fold domain

0

20

40

60

80

100

0 10 20 30 40 50 60

Time (min)

Prod

uct (

%) - hRMI1

+hRMI1

- -RMI1 +RMI1

Time-course

hRMI1 strongly stimulates BLM/hTOPOIIIα-dependent dissolution

hRMI1 STIMULATES DISSOLUTION VIA TOPO IIIα

0

20

40

60

80

100

0 2 4 6 8 10

[TOPO IIIα]

Prod

uct (

%)

[BLM] nM

hRMI1hTOPO IIIα

BLM

- hRMI1+hRMI1

1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 7 8 9 10 11 12 13 14

- hRMI1+hRMI1

Prod

uct (

%)

20 40 0.5 1 1.5

nM

2

5nM 5nM

125nM 125nM

166nM 166nM

020406080

100

0

AB

AB

hTOPO IIIα (nM)90452312631.50

18nM hTOPO IIIα 54nM hTOPO IIIα

hRMI1

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 2120

RMI1 recruits hTOPOIIIα to the double Holliday junction

Summary and Additional Data

BLM and hTOPO IIIα catalyze a novel mechanism to resolve recombinationintermediates that does not involve RuvC-like endonuclease cleavage (Double junction dissolution)

Double junction dissolution is markedly enhanced by hRMI1 via a stimulationof topo IIIα. hRMI1 binds directly to topo IIIα, and appears to promote its loading to the substrate

Double junction dissolution is highly specific for BLM, but requires the action of any type IA topoisomerase

The HRDC domain of BLM (and RecQ) constitutes a DNA structure-specificrecognition motif

The HRDC domain of BLM (and Lys-1270) is required for efficient catalysis of dissolution

Alternative template usage during homologous recombination

sister chromatid exchange loss of heterozygosity chromosome

instability

Resolution into crossover products

BLM/RMI1/hTOPO IIIα

CANCER