What’s new in the biology of CML?

John Goldman Imperial College London

Newcastle, 1st March 2013

Hammersmith Hospital London

Molecular features of BCR-ABL1

BCR-ABL1 as a clinical target

Signal transduction pathways

Signal transduction molecules as clinical targets

Stem cells as clinical targets

Some outstanding biological issues in CML

Molecular features of BCR-ABL1

BCR-ABL1 as a clinical target

Signal transduction pathways

Signal transduction molecules as clinical targets

Stem cells as clinical targets

Some outstanding biological issues in CML

BCR-ABL1 activates a myriad of signaling pathwaysBCR-ABL1 activates a myriad of signaling pathways

Goldman and Melo, NEJM 2003; 349: 1451, modified by van Etten, 2010

Tyr177

STAT1

STAT5P

LYNP

JAK2P

AHI1

BCL6

mTOR

Smo

GLI-1

GSK3

eIF4E

CatP

BCLX

Molecular features of BCR-ABL1

BCR-ABL1 as a clinical target

Signal transduction pathways

Signal transduction molecules as clinical targets

Stem cells as clinical targets

Some outstanding biological issues in CML

BCR-ABL1 activates a myriad of signaling pathwaysBCR-ABL1 activates a myriad of signaling pathways

Goldman and Melo, NEJM 2003; 349: 1451, modified by van Etten, 2010

Tyr177

STAT1

STAT5P

LYNP

JAK2P

AHI1

BCL6

mTOR

Smo

GLI-1

GSK3

eIF4E

CatP

BCLX

Approaches to increasing the efficacy of TKI

Exploitable differences

EpigeneticsHLA class II

Cell cycle regulationChemokinesPML/PTEN

PP2AFoxOBCL6

CytokinesCXCR4

WntHedgehog

Autophagy Alox 5/15

Immune basedGenomic instability

ROS

Actual approaches

HDAC-inh.ArsenicFTY720

TGFinh.RI-BP1

Jak2-inh.Plerixafor

GSK3inh.Smo inhibitors0H-chloroquine

ZileutonImmunotherapy

PARP-inh.

Approaches to increasing the efficacy of TKI

Exploitable differences

EpigeneticsHLA class II

Cell cycle regulationChemokinesPML/PTEN

PP2AFoxOBCL6

CytokinesCXCR4

WntHedgehog

Autophagy Alox 5/15

Immune basedGenomic instability

ROS

Actual approaches

HDAC-inh.&Demeth. agents

ArsenicFTY720

TGFinh.RI-BP1

Jak2-inh.Plerixafor

GSK3inh.Smo inhibitors0H-chloroquine

ZileutonImmunotherapy

PARP-inh.

Capacity of catenin to enhance LSC self renewal

Capacity of catenin to enhance LSC self renewal

Possible role of PI3 kinase and AKT in signal transduction

Possible role of PI3 kinase and AKT in signal transduction

Role of protein phosphatase 2A in CML

Cancerous inhibitor of PP2A (CIP2A) predicts progression to BT

Lucas et al, Blood 2011; 117: 6660-6668

PP2A is a tumor suppressor that is inactivated by BCR-ABL1 signaling,

CIP2A also renders protein phosphatase 2A (PP2A) inactive

Patients with high levels of CIP2A progress to BT, whereas patients with lower levels do not, and

CIP2A may act through altered phosphorylation of MYC on S62

SET

JAK2

IM

PP2A

c-MycSer62

BCR-ABL1

CIP2A

Lucas et al, Blood 2011; 117: 6660-6668

Diagnosis Diagnosis Diagnosis

CCyR No CCyR BC

CIP2A at diagnosis as a predictive marker for response to TKI

25

20

15

10

5

0

CIP

2A

(M

FI)

Probability of disease progression stratified by CIP2A level at diagnosis

Patients with a high diagnostic CIP2A protein level have 100% probability of progressing to blast crisis (p=<0.0001).

Normal

A B

Low CIP2AHigh CIP2A

C

C C R e N o C C R e B C

p=<0.001

p=0.007

Diagnosis 12 months

C C R e N o C C R e B C

Diagnosis 12 months Diagnosis 12 months

MNC CD34+ cells

Progression-free survival

P=0.017

Lucas CM et al. Blood 2011

Low

High

Sonic hedgehog (Shh or HH) in CML

Activation of the Sonic hedgehog pathway in chronic myeloid leukemia

The levels of Shh, Smo and Gli are increased in chronic phase and considerably more increased in

advanced phases of CML

Bing Long et al, J Exper Clin Cancer Research 2011; 30:8

Expression of Shh and its receptors in different phases of CML

Shh Patched Smo Gli1

Sonic hedgehog (Shh or HH) in CML

SHH + PTCH1

SMO

GLI1GSK3

Sufu

P

Target Genes(e.g. PTCH1, GLI1, Myc)

sFRP-1

-catenin

Target Genes(e.g. cyclinD1, CD44)

Wnt + Frizzled?

-catenin + TCFGLI1+GLI3

Proposed cross-talk between Shh and WNT pathways in CML

Sabutoclax, a pan BCL2 inhibitor renders bone marrow resident leukemia stem cell

sensitive to TKI

Goff et al, Cell Stem Cell, 2013; 12: 1-13

Peptide inhibition of BCL6 with retroinverso BCL6 peptide inhibitor (RI-BPI) compromises colony formation in CML and selectively eradicates CD34+ CD38- leukemia initiating cells

Hurtz et al. J Exper Med 2011; 208: 2016

BCL6-mediated repression of p53 is critical for CML LSC survival

BCL6-mediated repression of p53 is critical for CML LSC survival

Hurtz et al, J Exper Med 2011;11: 2163-2174

Inhibitory effect of a retro-inverso BCL6 peptide Inhibitor (RI-BPI)

ADAR1 promotes myeloid progenitor re-programming in CML

Jiang et al, PNAS 2013; 110: 1041

Isoform diversity can be generated by RNA editing by adenosine deaminase activation of RNA (ADAR) enzymes that regulate stem cell maintenance.

Jiang et al used serially transplantable blast crisis progenitors in a mouse model and showed BCR-ABL amplification and enhanced expression of an IFN-responsive ADAR1 p150 isoform

The authors suggest that techniques directed against ADAR1 might be useful in eradication of CML LSC

Nilotinib and MEK inhibitors induce synthetic lethality through paradoxical activation of RAF in drug-resistant CML

Imatinib, nilotinib and dasatinib all possess weak off-target activity against BRAF and CRAF in a RAS dependent manner. Because RAS is activated by BCR-ABL RAS activity may persist in the presence of TKIs. The authors have shown that nilotinib synergises with MEK inhibitors to kill drug resistant CML cells and block leukemia growth in mice

Packer et al, Cancer Cell 2011; 20: 715-727

Nilotinib and MEK inhibitors induce synthetic lethality through paradoxical activation of RAF in drug-resistant CML

PD = PD184352

Nilotinib and MEK inhibitors induce synthetic lethality through paradoxical activation of RAF in drug-resistant CML

Packer et al, Cancer Cell 2011; 20: 715-727

Can we target the quiescent stem cell without targeting as signal transduction

molecule?

12-prostglandin J3 was identified by screening publically

acquired databases of cyclopentenone prostaglandins.The agent alleviates the development of leukemia in a murine model system – both in mice infected with a Friend erythro-leukemia virus and mice expressing a BCR-ABL oncogene in their haematopoietic compartment

12-prostglandin J3, an omega 3 fatty acid-derived metabolite, selectively ablates leukemia stem cell in mice

Hegde et al, Blood 2011;118: 6909

12-prostaglandin J3, an omega 3 fatty acid-derived metabolite, selectively ablates

leukemia stem cell in mice

Hegde et al, Blood 2011;118: 6909

Possible role of constitutional factors in dictating responses to TKIs

A common polymorphism in the BIM gene contributes to intrinsic resistance to IM

Used massively parallel DNA sequencing of paired end ditags

Identified a novel deletion polymorphism in the BIM gene that is associated with deletion of exon 4 and the BH3 domain

12.3% of East Asian persons have this BIM deletion which is not seen in Africans or Caucasians

The BIM deletion inhibits imatinib-induced apoptosis in CML cells and so results in resistance to imatinib as well as to other TKIs

Ng, Tiong-Ong et al, Nat Med 2012;18: 521-528

Resistance to TKI results from deletion of exon 4 that contains BIMi2+/+ encoding BH3 transcripts that activate

apoptosis

Ng, Tiong-Ong et al, Nat Med 2012;18: 521-528

x

Ng, Tiong-Ong et al, Nat Med 2012;18: 521-528

Ng, Tiong-Ong et al, Nat Med 2012;18: 521-528

Association of a BIM deletion polymorphism with clinical resistance to imatinib in CML-CP

Have we neglected study of possible epigenetic factors in dictating

disease progression ?

AACR, Nature 2008; 454: 711

Chromosome

Transcription

Chromatin remodeller

Histones

Histone tails

Transcription

Non coding RNAs

Targeting epigenetic mechanisms in CML

Histone tail modifications DNA methylation

Chromatin remodeling

Non-coding RNAs

AACR, Nature 2008; 454: 711

Chromosome

Transcription

Chromatin remodeller

Histones

Histone tails

Transcription

Non coding RNAs

Targeting epigenetic mechanisms in CML

DNA methylation

Histone tail modifications Chromatin remodeling

Non-coding RNAs

Are there specific differences in the pattern of DNA methylation in CML and normal cells?

Does the pattern of methylation differ in different phases of CML?

Does the pattern of DNA methylation differ within chronic phase?

Does the pattern of DNA methylation differ in different sub-populations of Ph+ cells, eg CD34+ cells vs. GMP?

Questions that can now be addressed with epigenome wide studies to identify

aberrant DNA methylation

Are there specific differences in the pattern of DNA methylation in CML and normal

cells?

Bazeos et al. used Illumina 450 array based technology to study the methylation status CpG dinucleotides in DNA from CD34+ cells from 46 patients with CML-CP and normal controls

Results were compared with methylation patterns of 29 of the patients after they had achieved CCyR

DNA methylation analysis was clearly able to distinguish CML-CP, CML-CCyR and normal CD 34+ cells

Bazeos et al, poster presented at EHA-Amsterdam – June 2012

DIAGNOSIS NORMALS

CCyR

Unsupervised hierarchical clustering (Bazeos et al, 2013)

DNA methylation patterns in CMLH

eigh

t

Do we rely too much on measurement of BCR–ABL transcripts?

Residual disease: % BCR-ABL/ABL mRNA % BCR-ABL DNA

28.22

1.82 2.08

0.030.060.04

0.09

UND UND

2.228

3 6 12 15 18

Months

0.054 0.016 0.009 0.006 0.0050.003 0.002

0.001

21 27 36 42

Low Level Positivity of g-PCR When RT-PCR Is Negative

Mattarucchi E, et al. J Mol Diagn. 2009;11:482-487.

Molecular monitoring of residual in CML – RT-PCR compared with genomic PCR

DNA-based PCR may be more sensitive than RQ-PCR

Ross et al, Leukaemia 2010

BCR-ABL breakpoint cloning strategy

Alikian et al, 2013 paper submitted

DNA-based MRD (NGS)

Alikian et al, 2013 submitted

What about blastic transformation?

Pathways to blastic transformation

Adapted from: Perrotti D, et al. J Clin Invest 2010; 120;2254

One approach to improving the efficacy of TKI should be to add a signal transduction inhibitor, but the ideal agent as not yet been identified

It should be possible to identify agents that act directly on the LSC, regardless of whether the LSC is quiescent or active and regardless of whether it is or is not oncogene addicted

Conclusions (1)

The fundamental heterogeneity of CML remains largely unexplained. It could be due in part to constutional factors or to epigenetic factors. The pattern of DNA methylation of CML cells differs substantially from that of their normal counterparts.

A DNA based PCR might be more informative than a RNA based PCR, especially when discontinuation of TKI is being considered.

Conclusions (2)

Imperial College Non-Imperial

Jane Apperley Michael Deininger

Mary Alikian Ong Tiong

David Marin Tessa Holyoake

Gareth Gerrard Tom O’Hare

Alex Bazeos Nick Cross

Thanks especially to

15th INTERNATIONAL CONFERENCE ON CHRONIC MYELOID LEUKEMIA:

BIOLOGY AND THERAPY                                 

Estoril, PortugalSeptember 26-29, 2013

Chairs: J. M. Goldman, J. Cortes, T. P. Hughes