www.sciencetranslationalmedicine.org/cgi/content/full/4/148/148ra117/DC1

Supplementary Materials for

Targeted Disruption of the BCL9/ββββ-Catenin Complex Inhibits Oncogenic Wnt Signaling

Kohichi Takada, Di Zhu, Gregory H. Bird, Kumar Sukhdeo, Jian-Jun Zhao, Mala Mani,

Madeleine Lemieux, Daniel E. Carrasco, Jeremy Ryan, David Horst, Mariateresa Fulciniti, Nikhil C. Munshi, Wenqing Xu, Andrew L. Kung, Ramesh A. Shivdasani,

Loren D. Walensky,* Daniel Ruben Carrasco*

*To whom correspondence should be addressed. E-mail: Ruben_Carrasco@dfci.harvard.edu (D.R.C.); Loren_Walensky@dfci.harvard.edu (L.D.W.)

Published 22 August 2012, Sci. Transl. Med. 4, 148ra117 (2012)

DOI: 10.1126/scitranslmed.3003808

The PDF file includes:

Materials and Methods Fig. S1. BCL9 overexpression in a broad range of human tumor types. Fig. S2. Cellular uptake and stability of SAH-BCL9B. Fig. S3. Selective inhibition of BCL9/β-catenin–driven transcription by SAH-BCL9B. Fig. S4. SAH-BCL9B enhances the cytotoxic effect of conventional chemotherapeutic agents. Fig. S5. Effect of SAH-BCL9B in noncancer cells. Fig. S6. Histology of tumor tissue, colonic mucosa, and bone marrow of SAH-BCL9B–treated mice. Fig. S7. Proliferation and apoptosis of INA-6 cells. Fig. S8. Wnt transcriptional activity after SAH-BCL9B treatment.

Supplementary Material and Methods

Chromatin Immunoprecipitation Three micrograms of antibody was prebound for 8h to protein A and protein G Dynal magnetic beads (Dynal Biotech, Norway) and washed 5x with ice-cold PBS containing 5% BSA, and then added to the diluted chromatin for overnight immunoprecipitation using the following antibodies: TCF-4 (Upstate #05-511), mouse IgG2a isotype control (Sigma, M5409), and rabbit IgG (sc-2027, Santa Cruz). The magnetic bead-chromatin complexes were collected and washed 6x in RIPA buffer (50 mM HEPES [pH 7.6], 1 mM EDTA, 0.7% Na-deoxycholate, 1% NP-40, 0.5 M LiCl). DNA was eluted from the beads as previously described(54). Amplification was carried out with a PTC-200 programmable thermal controller (MJ Research) after an initial denaturation at 94°C for 5min, followed by 30 cycles of PCR using the following temperature and time profile: denaturation at 94°C for 0.5min, primer annealing at 59°C for 0.5min, primer extension at 72°C for 0.5min, and a final extension of 72°C for 10min. The PCR products were visualized by 2% gel electrophoresis. The following promoter primer sets were employed: (1) VEGF (55): gcgtgtctctggacagag-ttt; agcctcagcccttccaca; (2) VEGF upstream: gaggctatgccagctgtagg; ccctt-ttcctccaactctcc; (3) c-MYC:actcccccggctcggtccacaagc; cccaatttctcagccaggtttcag.

Patient Specimens and Cell Lines Bone marrow specimens were obtained from patients with MM in accordance with Dana-Farber Cancer Institute Review Board approval and informed consent performed in compliance with the Declaration of Helsinki. Primary CD138+ plasma cells were purified using magnetic beads as described (16). CRC primary tumor samples were obtained from the Brigham and Women’s Hospital in accordance with the policies of their Institutional Review Board. To generate sufficient CRC primary tumor cells for experimentation, the primary tumors were first expanded subcutaneously in NOD/SCID mice (Jackson Laboratory). After the tumors reached 2cm in diameter, mice were sacrificed according to institutional guidelines and subcutaneous tumor xenografts were minced with a scalpel and digested by incubation with collagenase IV (Worthington Biomedical Corporation) and 0.01% DNase I (Sigma-Aldrich) at 37°C for 30min, followed by additional mechanical disaggregation using a Stomacher device (Seward Laboratory Systems Inc.). Samples were filtered through a 70µm cell strainer and washed with PBS. Red blood cells were lysed using ACK lysing buffer (BioWhittaker, Lonza) and viable tumor cells were enriched by Ficoll-Paque gradient centrifugation (GE Healthcare). To purify viable tumor cells only, the samples were treated with APC conjugated anti-mouse H-2Kd (clone SF1-1.1.1, eBioscience), FITC-conjugated anti-human EpCAM antibodies (clone Ber-EP4, Dako), and Hoechst 33258 (Sigma-Aldrich), and then FACSAria flow sorting (BD Biosciences) was used to isolate the EpCAM-positive, H-2Kd-negative, and Hoechst-negative primary tumor cells. Cultured cell lines were maintained as previously described(19).

Quantitative Reverse Transcription-PCRRNA was extracted with TRIzol Reagent (Invitrogen) according to the manufacturer’s protocol. Total RNA (2 μg) was reverse transcribed (SuperScript VILO cDNA synthesis kit, Invitrogen) and qPCR was performed using an Applied Biosynthesis 7500 Real-time PCR system. Analysis of target genes was conducted in quadruplicate using POWER SYBR Green Master Mix (Applied Biosystems) as previously described(19). Transcripts levels were normalized to β-actin expression. These experiments were repeated three times. PCR primers are designed as below: FOXQ1: cgcggactttgcactttgaa; agctttaaggcacgtttgatggag. CDK4: atgttgtccggctgat -gga; caccagggttaccttgatctcc. Axin2: cggaaactgttgacagtggat; ggtgcaaagacatagcc -agaa. VEGF: catgaactttctgctgtcttgg; atgattctgccctcctcctt. LGR5: ctcccaggtctggtgt -gttg; gtgaagacgctgaggttgga. CMYC: tttttcgggtagtggaaaacc; gcagtagaaatacggctg-cac. CD44: tttgcattgcagtcaacagtc; tgagtccacttggctttctgt. CLDN2: cggtgtggctaagta -caggc; caaagctcacgatggtggtct. LEF-1: catcccttcctcattccttcaac; aggcttcctaaaaggt -ggtgg. Establishment of Single Cell Cultures Colo320 cells were subjected to infection for 72h with either 7xTdG or 7xFdG lentivirus, and the transduced and control non-transduced cells were trypsinized, washed, and then analyzed on a FACSaria flow sorter. Hoechst 33258 staining was used to exclude dead cells. Single GFP-positive Colo320-7xTdG cells were sorted into 96 well plates using stringent gating on forward/side scatter height and width to exclude doublets. The presence of a single cell per well was confirmed microscopically after sorting and then single cell cultures were expanded for subsequent use. Fluorescence Microscopy Cells were prepared using a cytocentrifuge (Thermo Shandon) and fixed as previously described(16, 19). Anti-β-catenin and rhodamine-conjugated secondary antibodies (5µg/ml; Southern Biotechnology) were employed. Images were obtained using a BioRad Radiance 2000 laser and Leica TCS SP5 scanning confocal microscopy. Lentiviral Vectors A lentiviral reporter vector containing seven TCF/LEF-1 binding motifs and a minimal promoter driving destabilized GFP expression (7xTdG) was derived from the lentiviral vector TOP-dGFP, which contains three TCF/LEF-1 binding motifs(45). Two synthetic complementary oligonucleotides (IDT-DNA) with four TCF/LEF-1 binding motifs (GATCAAAGG) were designed to generate compatible overhanging ends for annealing to an Xba1 restriction site. The oligonucleotides were annealed by heating to 95°C and slow cooling to room temperature, followed by ligating into the Xba1-linearized TOP-dGFP vector, yielding 7xTdG. For construction of the control vector carrying seven FOP-sites (7xFdG), the 7xTOP cassette was removed from the 7xTdG vector by restriction digesting with Xma1 and Age1. A synthetic cassette carrying seven FOP sites (GGCCAAAGG)

but otherwise identical to the removed 7xTOP cassette was inserted, yielding 7xFdG. BCL9 shRNA and control shRNA lentiviral vectors were generated as reported(56). Lentivirus Production and Infection HEK293T cells were plated in 10cm tissue culture dishes and co-transfected with 10µg lentiviral vector (either 7xTdG or 7xFdG), 10µg pCMV-dR8.91 and 2µg pMD2.G (57) using 60µL LipoD293 (Signagen) according to the manufacturer’s protocol. The media was replaced after 12h with 30% FCS containing DMEM (Gibco) and conditioned for 36h. Conditioned medium was then filtered through 0.45µm syringe filters (Millipore), mixed 1:1 with fresh DMEM, and then directly used for infection of cultured Colo320 (ATCC) cells. Polybrene (Sigma) was added to a final concentration of 8µg/mL to enhance the efficiency of infection. Lentivirus shRNA infections to knockdown BCL9 expression were performed as described previously (19). Briefly, recombinant BCL9 shRNA and control lentiviruses were produced by transient transfection of 293T cells. Colo320 were transduced with virus supernatant containing polybrene, and GFP-expressing cells sorted by FACS. VEGF ELISA assay VEGF ELISA was performed as previously described (19). Briefly, cells (1x106) were treated with vehicle and SAH-BCL9 peptides (5 μM) for 24 h. VEGF levels in the supernatant were then measured according to the manufacturer’s ELISA protocol (DuoSet, R&D Systems). Protease Resistance of Peptides Protease reaction samples contained 5 μL peptide in DMSO (1 mM stock) and 195 μL of buffer consisting of 10 mM Tris-HCl buffer pH 7.4 with 1 mM CaCl2 and chymotrypsin (0.5 ngμL). Chymotryptic peptide fragments were analyzed by liquid chromatography/ mass spectrometry (Agilent 1200).

Nuclear/Cytosol Fractionation Kit 2 x 106 cells are performed to separate nuclear extract and cytoplasmic fraction according to protocol of BioVision Nuclear/Cytosol Fractionation Kit, #K266-25.

BCL9 high BCL9 low

12/23 (52.2 %) 11/23 (47.8 %)

16/22 (72.7 %)6/22 (27.3 %)

15/32 (46.9 %) 17/32 (53.1 %)

24/29 (82.8 %)

24/32 (75.0 %) 8/32 (25.0 %)

5/29 (17.2 %)

colo

nbr

east

lung

liver

ovar

ian

Anti-BCL9 Ab Anti-BCL9 Ab + BCL9 peptide

12/23 (52.2 %) 11/23 (47.8 %)

16/22 (72.7 %)6/22 (27.3 %)

15/32 (46.9 %) 17/32 (53.1 %)

24/29 (82.8 %)

24/32 (75.0 %) 8/32 (25.0 %)

5/29 (17.2 %)

A

B

SUPPLEMENTARY FIGURES

Fig. S1. BCL9 overexpression in a broad range of human tumor types. (A) Immunohistochemical studies performed on tissue microarrays from colon (n=23), breast (n=22), lung (n=32), liver (n=29), and ovarian (n=32) carcinomas revealed high levels of BCL9 expression in a wide variety of tumors. Representative cases of tumors with high or low level of BCL9 immunostaining are shown. Scale bar: 50�m. (B) Blocking experiments using the immunizing BCL9 peptide (Abcam) were performed on colorectal cancer specimens according to the manufacturer’s protocol and documented BCL9 antibody specificity. Scale bar: 50�m.

HoechstFITCBright image elciheV

9LC

B-H

AS

B

Cytosol Nucleus

Lamin B

FITC

Actin

Vehicle

SAH-BCL9 B

Vehicle

SAH-BCL9 BB

A

37014

Compound Half-Life(min) (fold)

Enhancement

126SAH-BCL9B

TAT-BCL9 HD2

C

D

E

F

SAH-BCL9TAT-BCL9HD2

0 1 3 6 12 0 1 3 6 12 0 1 3 6 12 (hr) SAH-BCL9B

MUT

SA

H-B

CL9

TA

T-B

CL9

HD

2S

AH

-BC

L9B

MU

T

DAPI FITC

FITC

Actin

GRKKRRQRRRPQLSQEQLEHRERSLQTLRDIQRMLF

LSQEQLEHRERSLXTLRXIQRBLF

LSQEQLEHEERSLXTLRXIQRBLFSAH-BCL9

SAH-BCL9 B

MUT

TAT-BCL9HD2

490 35SAH-BCL9 MUT

0 100 200 300 400 5000

0.5

1.0

Time (Min)

Frac

tion

Inta

ct

DAPI FITC

DLD1 Colo320

Fig. S2. Cellular uptake and stability of SAH-BCL9B. Cellular uptake of vehicle and FITC-SAH-BCL9B was monitored by live cell microscopy (A), and cellular fractionation and immunoblot (B) analysis in DLD1 cells. (C) Amino acid sequence of TAT-BCL9HD2, SAH-BCL9B and SAH-BCL9MUT peptides. (D) Cellular uptake of FITC-TAT-BCL9HD2, FITC-SAH-BCL9B and FITC-SAH-BCL9MUT in Colo320 cells as evaluated by immunoblot analysis of total cell lysates. (E) Cellular uptake of FITC-TAT-BCL9HD2, FITC-SAH-BCL9B and FITC-SAH-BCL9MUT peptides in DLD1 (left) and Colo320 (right) cell lines as evaluated by confocal microscopy after 2 hr incubation with 1�m peptide concentration. (F) Comparative resistance of FITC-TAT-BCL9HD2, FITC-SAH-BCL9B and FITC-SAH-BCL9MUT peptides to chymotrypsin. Chymotryptic peptide fragments were analyzed by liquid chromatography/mass spectrometry (Agilent 1200).

βCoomassie anti- catenin150

37

50

75

100

Lane: 1 2 1 2

Coomassie

1 2

150

37

50

75100

25

A

B

FITC

Actin

Colo320 MM1S

SAH-B

CL9B

SAH-BCL9 MUT

Vehicl

e

4°C 37°C25°C

kDa kDa

FITC

Actin

β-catenin

IκB-α

actin

FITC

actinIκB-α

β-catenin

IP: F

ITC

TCL

IP:

β-ca

teni

nTC

L

E-cadherin

β-catenin

β-catenin

FITC

E-cadherin

FITC

FITC

C

SAH-B

CL9B

SAH-BCL9 MUT

Vehicl

e

SAH-B

CL9B

SAH-BCL9 MUT

Vehicl

e

SAH-B

CL9B

SAH-BCL9 MUT

Vehicl

e

Fig. S3. Selective inhibition of BCL9/�-catenin-driven transcription by SAH-BCL9B. (A) Full-length GST-tagged h�-catenin was expressed and then purified to homogeneity, as evaluated by Coomassie staining and anti-�-catenin immunoblot. GST-�-catenin input: Lane 1, 0.2 �g and Lane 2, 0.4 �g (left). His-tagged hBCL9 (residues 243-469) containing the HD2 domain (Lane 1) and full-length GST-tagged h�-catenin (Lane 2) were produced and purified for use in GST-pull down assays (right). (B) FITC-SAH-BCL9B and FITC-SAH-BCL9MUT peptides exhibit equivalent temperature-dependent uptake in Colo320 cells (top), and cellular uptake in representative colorectal and myeloma cell lines (bottom). (C) �-catenin targeting by FITC-SAH-BCL9B is selective for disruption of the BCL9/�-catenin complex and does not affect anti-�-catenin co-immunoprecipitation of E-cadherin from MCF7 cell lysates (top). Treatment of MCF7 cells with FITC-SAH-BCL9B followed by anti-FITC pull down performed on cellular lysates revealed the selective interaction between FITC-SAH-BCL9B and �-catenin, and no co-immunoprecipitation of unrelated proteins such as I�B� and actin (bottom). MCF7 cells were employed in these assays as they contain readily detectable levels of E-cadherin protein for monitoring the �-catenin/E-cadherin interaction. TCL: total cellular lysate.

(μM)

SAH-BCL9 B

SAH-BCL9 MUT

Vehicl

e

**

0

20

40

60

80

100

120

Cel

l gro

wth

(% C

ontro

l)

5 10 20 5 10 20

*

A B

BCL9

B9L

β-catenin

actin

β

Sample 1 Sample 2 Sample 3

Vehicle

SAH-BCL9B

0

20

40

60

80

100

120

0 10 20 30(mM)5-FU

SAH-BCL9MUT

Cel

l gro

wth

(%

Con

trol)

** **

***

Vehicle

SAH-BCL9BSAH-BCL9MUT

**** **

*0 100 200 300

(nM)Dox

D

E

Colo32

0

DLD1SW

48HCT1

16P

HCT116

D

HCT116

K

Ls17

4T

MM1S

OPM1H92

9IN

A-6

BCL9

B9L

-catenin

actin

C

Fig. S4. SAH-BCL9B enhances the cytotoxic effect of conventional chemotherapeutic agents. (A) SAH-BCL9B dose-responsively inhibited the growth of Colo320 cells. *p < 0.01. Expression of BCL9, B9L and �-catenin in CRC (B) and MM (C) cell lines as evaluated by immunoblot. (D) Immunohistochemical analysis of BCL9 expression (brown) in three different cases of CRC (top) and MM (bottom) primary tumors used in the cellular proliferation analyses. Insets highlight expression of BCL9 and its nuclear localization in the tumor cells. Scale bar: 50�m. (E) Colo320 (left) and MM1S (right) cells were co-cultured with either vehicle, SAH-BCL9B, or SAH-BCL9MUT and increasing concentrations of 5-flurouracil (5-FU) or doxorubicin (Dox), and evaluated for cellular proliferation using 3H-thymidine incorporation. SAH-BCL9B but not SAH-BCL9MUT or vehicle significantly enhanced the anti-tumor activity of the conventional agents.

Colo32

0

PBMC-1

PBMC-2

WS1

BJ Colo32

0

BCL9

�-catenin

ActinSkin

fibro

blast

#1

WS

1C

olo3

20

DAPI BCL9 ������ Merge

Actin

Lamin B

BCL9

������

Luci

fera

se/R

enill

a ra

tio

DA

E

C

������

Actin

BCL9

Lamin B

Cyt

osol

Nuc

leus

180

100

5

0

SAH-BCL9 B

SAH-BCL9 MUT

Vehicl

e

0

0.2

0.4

0.6

0.8

1

1.2

ConA

Vehicl

e- + + +

Cell

gro

wth

(%

contr

ol)

B

Fig. S5. Effect of SAH-BCL9B in noncancer cells. (A) Immunoblot analysis of the expression of BCL9, B9L and �-catenin in normal peripheral blood mononuclear cells (PBMC-1 and PBMC-2), normal skin fibroblasts (WS1 and BJ), and Colo320 cells. (B) Effect of SAH-BCL9B treatment on the growth of PBMC-1 cells as evaluated by [3H]Thymidine incorporation after stimulation with Concanavalin A for 3 days. Error bars are mean +/-s.d. for experiments performed in quadruplicate. (C) Colo320 and WS1 cells were transfected with TOP-FLASH and FOP-FLASH, incubated with vehicle (red) or SAH-BCL9B peptide (10�M), and assayed for luciferase activity, which was normalized to Renilla luciferase control, and TOP-FLASH luciferase activity was normalized to that of FOP-FLASH luciferase activity. Error bars are mean +/- s.d. for assays performed in triplicate. (D) Cellular fractionation and immunoblot analyses of BCL9 and �-catenin expression in Colo320 (top) and WS1 (bottom) cells. Lamin B and Actin were used as markers of nuclear and cytoplasmic fractions, respectively. (E) Confocal immunofluorescence analysis of BCL9 (red) and �-catenin (green) expression in WS1 (top) and Colo320 (bottom) cells.

Vehicle

Sam

ple

1Sa

mpl

e 2

Sam

ple

3

SAH-BCL9B SAH-BCL9 MUT

0

10

20

30

40

50

60

70

*

SAH-BCL9 B

SAH-BCL9 MUT

Vehicl

e

TUN

EL

Pos

itive

Per

cent

age

(%)

A

Vehicle SAH-BCL9B SAH-BCL9 MUTB

Fig. S6. Histology of tumor tissue, colonic mucosa and bone marrow of SAH-BCL9B-treated mice. (A) Tumor tissue sections from NOD/SCID mice bearing intraperitoneal Colo320 cells were evaluated for apoptosis induction using the TUNEL assay (brown). SAH-BCL9B, but not vehicle or SAH-BCL9MUT, notably increased TUNEL positivity. Three representative 40 X power fields are shown, including quantitation of TUNEL positivity in 3 high power fields. * p< 0.001. Scale bar: 50�m. (B) H&E staining of colonic mucosa (top) and bone marrow (bottom) tissues isolated from experimental mice injected with colo320 cells and treated with vehicle or SAH-BCL9-peptides showed no evidence of toxicity across all histologic specimens. Scale bar: 125�m.

,

120

100

80

60

40

20

0C

ell g

row

th(%

Con

trol)

Mock

EdTP

*

B

A

Sam

ple

1Sa

mpl

e 2

Sam

ple

3

0

5

10

15

20

25

TUN

EL

Pos

itive

Per

cent

age

(%)

*

SAH-BCL9 B

SAH-BCL9 MUT

Vehicl

e

Vehicle SAH-BCL9B SAH-BCL9 MUT

Fig. S7. Proliferation and apoptosis of INA-6 cells. (A) INA-6 cells were lentivirally transduced with empty vector (Mock) or a vector expressing a dominant negative form of TCF4 (EdTP) and proliferation was measured by 3H-thymidine incorporation. (B) Tumor tissue sections from NOD/SCID mice bone chips bearing INA-6 cells were evaluated for apoptosis induction using the TUNEL assay (brown). SAH-BCL9B, but not vehicle or SAH-BCL9MUT, notably increased TUNEL positivity. Three representative 40 X power fields are shown, including quantitation of TUNEL positivity in 3 high power fields. * p< 0.001. Scale bar: 50�m.

SAH-BCL9 ( - )

Proliferation

Migration, Invasion

IntratumoralAngiogenesis

SAH-BCL9 ( + )

Nucleus

BCL9

Cancer Cell

c-Myccyclin D1VEGFCD44

Nucleus

TCF4

�-catenin

Cancer Cell

c-Myccyclin D1VEGFCD44

BBCBCLBCLCLBBCLB -9-9-999BBBBCCCCLCLB -999BBBBCLCCLBCLL 99BBBCL 9BCL9

SAH-BCL9

TCF4

�-catenin

TBS TBS

Fig. S8. Wnt transcriptional activity after SAH-BCL9B treatment. Deregulated Wnt signaling underlies the pathogenesis of a broad range of human cancers yet the development of targeted therapies to disrupt the pathway has remained a challenge. �-catenin is a central effector of the canonical Wnt pathway, activating the expression of genes such as c-Myc, cyclin D1, VEGF, and CD44 that are involved in cell proliferation, migration, and angiogenesis. BCL9 is an important co-activator for �-catenin-mediated transcription, and is highly expressed in tumors but not in the cells of origin, presenting an opportunity to selectively inhibit pathologic �-catenin activity. Guided by the structure of the BCL9/�-catenin complex, we generated Stabilized Alpha-Helices of BCL9 (SAH-BCL9) to block Wnt signaling in cancer through targeted disruption of BCL9/B9L/�-catenin complexes. SAH-BCL9 reduces Wnt transcriptional activity and the expression of Wnt/�-catenin transcriptional targets, impeding tumor cell proliferation, migration, invasion, and angiogenesis in vitro and in vivo. TBS; TCF4 binding site.