1

SI Appendix April 12, 2014

A mechanistic rationale for targeting the unfolded protein response in pre-B

acute lymphoblastic leukemia

Behzad Kharabi Masouleh1,3, Huimin Geng1, Christian Hurtz1, Lai N. Chan1 Aaron C Logan2,

Mi Sook Chang4, Chuanxin Huang5, , Srividya Swaminathan1, Haibo Sun6, Valeria Cazzaniga1,7,

Giovanni Cazzaniga7, Elisabeth Paietta8, Ari Melnick5, H. Phillip Koeffler6, Markus Müschen1,2

Supplemental Material and Methods

Figures S1-S10

Tables S1-S5

2

SUPPLEMENTAL MATERIAL AND METHODS

Primary human samples and human cell lines

We obtained primary cases (Table S1) in compliance with the IRB of the University of California

San Francisco. The human Ph+ ALL, lymphoma or multiple myeloma cell lines were obtained

from DSMZ, Braunschweig, Germany. We maintained human leukemia cells in Roswell Park

Memorial Institute medium (RPMI-1640, Invitrogen, Carlsbad, CA) with GlutaMAX containing

20% fetal bovine serum, 100 IU ml−1 penicillin and 100 μg ml−1 streptomycin at 37°C in a

humidified incubator with 5% CO2. We cultured all primary human ALL cells on OP9 stromal

cells.

Extraction of bone marrow cells from mice

Bone marrow cells were extracted from young age-matched mice with different genetic

backgrounds (Table S2). We obtained the bone marrow cells by flushing cavities of femur and

tibia with PBS. After filtration through a 70 μm filter and depletion of erythrocytes using a lysis

buffer (BD PharmLyse, BD Biosciences), washed cells were either frozen for storage or subjected

to further experiments.

Mouse model of human Ph+ ALL

We collected bone marrow cells from the above mentioned mice and retrovirally transformed them

using BCR–ABL1 or NRASG12V in the presence of 10 ng ml−1 interleukin-7 (Peprotech) in

RetroNectin- (Takara) coated Petri dishes as described below. We maintained all BCR–ABL1-

transformed ALL cells derived from bone marrow of mice in Iscove’s modified Dulbecco’s

medium (IMDM, Invitrogen) with GlutaMAX containing 20% fetal bovine serum, 100 IU

ml−1 penicillin, 100 μg ml−1 streptomycin and 50 µM 2-mercaptoethanol. BCR–ABL1 or NRAS-

G12V-transformed ALL cells were propagated only for short periods of time and usually not longer

than for 2 months to avoid acquisition of additional genetic lesions during long-term cell culture.

BCR-ABL1 transformed pre-B cells were transduced with different empty vector controls (EV)

and either (4-OHT)-inducible or non-inducible Cre and deleted by Cre-mediated activation. A

detailed overview of the used plasmids is shown in Table S5, For in vivo experiments ALL cells

were then labeled with lentiviral firefly luciferase (D.B. Kohn, University of California Los

Angeles, Los Angeles, CA) and injected via tail into sublethally irradiated (250 cGy) NOD/SCID

recipient mice. Engraftment was monitored using luciferase bioimaging (VIS 100

bioluminescence/optical imaging system; Xenogen). D-Luciferin (Xenogen) dissolved in PBS was

injected intraperitoneally at a dose of 2.5 mg per mouse 15 min before measuring the light

3

emission. Seven mice per group were injected via tail vein injection. When a mouse became

terminally sick, it was sacrificed and bone marrow and spleen were collected for flow cytometry

analysis. All mouse experiments were subject to approval by the Children’s Hospital Los Angeles

Institutional Animal Care and Use Committee.

Retroviral transduction

We performed transfections of retroviral constructs and their corresponding empty vector controls

(Table S5) using Lipofectamine 2000 (Invitrogen) with Opti-MEM media (Invitrogen). We

produced retroviral supernatants to infect murine cells by cotransfecting 293FT cells with the

plasmids pHIT60 (gag-pol) and pHIT123 (ecotropic env; provided by D.B. Kohn, University of

California, Los Angeles, Los Angeles, CA). To infect human cells we replaced pHIT123 with

pHIT456 (amphotropic env; provided by D.B. Kohn, University of California, Los Angeles, Los

Angeles, CA). Cultivation was performed in high glucose DMEM (Invitrogen) with GlutaMAX

containing 10% fetal bovine serum, 100 IU ml−1 penicillin, 100 µg ml−1 streptomycin, 25 mM

Hepes, 1 mM sodium pyruvate, and 0.1 mM of nonessential amino acids. We replaced regular

media after 16 h with growth media containing 10 mM sodium butyrate. After 8 h of incubation,

the media was changed back to regular growth media. 24 h later, the virus supernatants were

harvested, filtered through a 0.45 µm filter, and loaded by centrifugation (2,000 g for 90 min at

32°C) two times on 50 µg ml−1 RetroNectin-coated non-tissue 6-well plates. 2–3 × 106 pre–B cells

were transduced per well by centrifugation at 500 g for 30 min and maintained overnight at 37°C

with 5% CO2 before transferring into culture flasks.

Senescence-associated β-galactosidase assay

Senescence-associated β-galactosidase activity was performed on cytospin preparations as

described (1). Briefly, a fixative solution (0.25% glutaraldehyde, 2% paraformaldehyde in PBS

pH 5.5 for mouse cells) was freshly generated. To this end, 1 g paraformaldehyde was dissolved

in 50 ml PBS at pH 5.5 by heating followed by addition of 250 μl of a 50% stock glutaraldehyde

solution. 1× X-gal staining solution (10 ml) was prepared as follows: 9.3 ml PBS/MgCl2, 0.5 ml

20× KC solution (that is, 820 mg K3Fe(CN)6 and 1,050 mg K4Fe(CN)6 × 3H2O in 25 ml PBS) and

0.25 ml 40× X-gal (that is, 40 mg 5-bromo-4-chloro-3-indolyl β-D-galactoside per milliliter

of N,N-dimethylformamide) solutions were mixed. For BCR–ABL1-transformed ALL cells,

100,000 cells per cytospin were used (700 r.p.m., 8 min). The fixative solution was pipetted onto

cytospins and incubated for 10 min at room temperature, then washed twice for 5 min in

4

PBS/MgCl2. Cytospin preparations were submerged in 1× X-gal solution, incubated overnight at

37 °C in a humidified chamber and washed twice in PBS.

Single-locus quantitative ChIP, ChIP-on-chip and ChIP-seq analysis

Cells were cross-linked with 1% formaldehyde for 10 minutes at room temperature, lysed, and

DNA was sonicated to a fragment size of 150-350bp. After sonication, immunoprecipitations were

performed with the following antibodies: anti-BCL6 (N3, Santa Cruz), anti-BACH2, or rabbit

control IgG (ChIP-grade, Abcam). DNA fragments enriched by ChIP were identified by

quantitatively PCR (ChIP-QPCR) using the Fast SYBR green kit. The GAPDH gene, which is not

a BCL6/BACH2 target, was used as negative control. The enrichment for each antibody at

examined loci is expressed as percentage of input. Primers are listed in Table S3. ChIP-seq

libraries were prepared using Illumina ChIP-seq Library preparation Kit according to the

manufacturer’s instructions with 10ng of purified ChIP DNA and sequenced for 36 cycles using

Illumina GAIIx sequencer. Peak calling and downstream analysis was performed using the

ChIPseeqer package (2). The ChIPseq data have been uploaded to GEO under accession number

GSE44420. ChIP-on-Chip was performed in three Ph+ ALL cell lines in duplicates as described in

(1) (GSE24404). Briefly, to identify target genes of BCL6 from each array, the log-ratio between

the probe intensities of the ChIP product and input was computed by taking moving windows of

three neighboring probes for the average log-ratio and determining the maximum value of those

windows as the signal for each probeset. Random permutation probe log-ratios were used as

background control. The peak cutoff was established as higher than the 99th percentile of the

24,175 log-ratio values generated from random permutation probes on the array. A locus with

maximum moving average above the cut-offs in both replicates, or in one replicate but the Pearson

correlation coefficient of the probes signal of the promoter between the two replicates higher than

0.8 (a rescue procedure), is considered a binding target.

Gene expression - microarray and RNAseq analysis

Microarray. Total RNA from cells used for microarray was isolated by RNeasy purification.

Biotinylated cRNA was generated and fragmented according to the Affymetrix protocol and

hybridized to mouse 430A 2.0 Array (Affymetrix). After scanning (GeneChip Scanner 3000 7G;

Affymetrix), the generated .cel files were imported to Expression Console software (Affymetrix)

and processed using the RMA average algorithm for normalization and summarization. RNAseq.

Total RNA were extracted and subjected to high-throughput sequencing. Illumina multiplexing

library construction, HiSeq 2000 SR50 sequencing, and data generation were performed by

5

Epigenomics Core Facility at Weill Cornell Medical College (New York, NY) according to

manufacturer's instructions. Raw image data were converted into base calls and fastq files via the

Illumina pipeline CASAVA version 1.8 with default parameters. All 50-bp-long reads were

mapped to the reference mouse genome sequence, mm9, using Tophat with the default parameters

(Trapnell et al., 2009). The mRNA expression level for each gene was represented as log2 RPKM

(reads per kilobase per million reads), called by Seqgene (Deng X, BMC Bioinformatics 2011).

Patient outcome and gene expression microarray data

Gene expression microarray and patient outcome data were obtained from the GEO database

accession numbers GSE5314 (3) and GSE34941 (Geng et al., 2012) of the Eastern Cooperative

Oncology Group (ECOG) Clinical Trial E2993 for adult B-ALL, from GSE11877 and GSE28460

(4) of the Children’s Oncology Group (COG) Clinical Trial P9906 for children B-ALL (the

National Cancer Institute TARGET Data Matrix, http://target

nci.nih.gov/dataMatrix/TARGET_DataMatrix.html), from the St. Jude Research Hospital

pediatric ALL clinical trial (raw data can be downloaded from

http://www.stjuderesearch.org/site/data/ALL3/). The microarray raw data were normalized using

the RMA method with Expression ConsoleTM software HG-U133 A and B (St Jude data, n=132)

or NimbleScan software (version 2.5, Roche NimbleGen, Madison, WI) for the NimbleGen arrays

HG18 60mer expression 385K platform (ECOG data). Expression level of a gene in a sample was

determined by the average of expression values from multiple probe sets on the array representing

this gene. The P-values of differential expression of a gene between BCR-ABL1 ALL and normal

pre-B samples and other ALL subtypes were determined by two-sided Wilcoxon test. The end

points of the clinical data include overall survival (OS), relapse-free survival (RFS), risk, time to

relapse, and expression levels in samples at the time of relapse vs. expression levels at diagnosis.

Array-based methylation analysis using HELP

The HELP (HpaII tiny fragment enrichment by ligation mediated PCR) assay was performed as

previously published (Geng et al., 2012). One microgram of high molecular weight DNA was

digested overnight with isoschizomer enzymes HpaII and MspI (NEB, Ipswich, MA). DNA

fragments were purified using phenol/chloroform, resuspended in 10mM Tris-HCl pH 8.0, and

used immediately to set up the ligation reaction with MspI/HpaII-compatible adapters and T4 DNA

ligase. Ligation-mediated PCR was performed with enrichment for the 200 to 2000 base pair (bp)

products and then submitted for labeling and hybridization onto a human HG_17 promoter custom-

designed oligonucleotide array covering 25,626 HpaII amplifiable fragments within the promoters

6

of the genes. HELP data analysis was performed as described previously (5), using R software and

Bioconductor package (http://www.bioconductor.org/). Basically, raw data (.pair) files were

generated using NimbleScan software. Signal intensities at each HpaII amplifiable fragment were

calculated as a robust (25% trimmed) mean of their component probe-level signal intensities. Any

fragments found within the level of background MspI signal intensity, measured as 2.5 mean-

absolute- deviation (MAD) above the median of random probe signals, were considered as “failed”

probes and removed. A median normalization was performed on each array by subtracting the

median log-ratio (HpaII/MspI) of that array (resulting in median log-ratio of 0 for each array).

References:

1. Duy C et al. (2011) BCL6 enables Ph+ acute lymphoblastic leukaemia cells to survive BCR-

ABL1 kinase inhibition. Nature 473:384–388.

2. Giannopoulou EG, Elemento O (2011) An integrated ChIP-seq analysis platform with

customizable workflows. BMC Bioinformatics 12:277.

3. Juric D et al. (2007) Differential gene expression patterns and interaction networks in BCR-

ABL-positive and -negative adult acute lymphoblastic leukemias. J Clin Oncol Off J Am Soc

Clin Oncol 25:1341–1349.

4. Hogan LE et al. (2011) Integrated genomic analysis of relapsed childhood acute

lymphoblastic leukemia reveals therapeutic strategies. Blood 118:5218–5226.

5. Geng H et al. (2012) Integrative epigenomic analysis identifies biomarkers and therapeutic

targets in adult B-acute lymphoblastic leukemia. Cancer Discov 2:1004–1023.

Fig. S1: Validation of Hspa5 (Ig heavy chain binding protein) deletion

Cre

b-actin

Hspa5

EV Cre Hspa5fl/fl ALLEV Cre

2 days 3 days

aHspa5fl/fl pre-B ALL +EVGFP

94 32 61 86

EV-GFP

CD

45.2

I II III IV

99 96 97 97

I II III IV

Cre-GFP

CD

45.2

Hspa5fl/fl pre-B ALL +CreGFP

b

Hspa5fl/fl pre-B ALL+EV +Cre Hspa5#1 2 3 4 5 6 7 8 9 #1 2 3 4 5 6 7 8 9 genotype

WTfloxed

deleted

c

Figure S1 legend: The protein levels of Hspa5 and Cre are measured by Western Blot analyses inHspa5fl/fl pre-B ALL cells with EV and Cre after two and three days of 4-OHT treatment using β-actinas loading control (a). Bone marrows and spleens from transplant recipient mice carrying Hspa5fl/fl

pre-B ALL with either CreGFP or EVGFP were stained for the surface markers CD45.1- and CD45.2+. Arepresentative analysis is shown (n=4) (b). A genomic PCR amplifying either the floxed allele(floxed), wildtype (WT) or after deletion (deleted) using a wildtype (C57BL6J), H2O and in vitrodeletion controls is shown for NOD/SCID mice (n=9) transplanted with Hspa5fl/fl pre-B ALL witheither CreGFP or EVGFP (c).

Cox6a2

Ccc

r7 m

RN

A le

vels

[%

Hprt

]

EV CreHspa5fl/fl EV Cre

60

50

40

30

20

10

0

Ccr7

1.6

Co

x6a2 m

RN

A [

% H

prt

]

EV CreHspa5fl/fl EV Cre

1.4

1

0.8

0.6

0.4

0.20

1.2

1.8

Ccn

g2 m

RN

A le

vels

[%

Hprt

]

EV CreHspa5fl/fl EV Cre

60

50

40

30

20

10

0

P=0.0007 P=0.001

Ccng2

Ccy

5rl m

RN

A le

vels

[%

Hprt

]

EV CreHspa5fl/fl EV Cre

10

8

6

4

2

0

Cy5rl

CD

22 m

RN

A le

vels

[%

Hprt

]

EV CreHspa5fl/fl EV Cre

20

15

10

5

0

P=0.0006 P=0.002

CD22

70P=0.001 P=0.01P=0.0003 P=0.003

P=0.0001 P=5.64 x 10-5

Fig, S2: Validation of microarray gene expression analysis of Hspa5-deletion

Hspa5fl/fl BCR-ABL1 ALL cells

b

a

c

e

d

Fig. S2 legend: The mRNA levels of Cox6a2, Ccng2, CD22, Ccr7 and Cy5rl were measuredby qRT-PCR in Hspa5fl/fl pre-B ALL cells transduced with empty vector control (EV) and 4-OHT-inducible retroviral Cre (Cre) after two days of 4-OHT treatment using Hprt as areference gene (n=2; a-e)

0

100

200

300

0

200

400

600

800

0

2000

4000

6000

8000

10000

12000

14000

XBP1

ERN1

HSPA5

0

500

1000

1500

2000PRDM1

mR

NA

leve

ls [

sig

na

l in

tensi

ty,

Aff

yme

trix

Ge

ne

Ch

IPa

rra

ys]

Xbp

1s

mR

NA

leve

ls [

% H

prt

]

EV µHC Ighm-/-0.00

0.05

0.10

0.15

0.20

P=0.0002Xbp1s

Xbp

1s m

RN

A leve

ls [

% H

prt

]

0

10

20

30

40

50

60

Fig. S3: Expression of plasma cell-specific unfolded protein response molecules ERN1, HSPA5, PRDM1 and XBP1 in normal pre-B cells and pre-B ALL

a

b

c

d

e

f

g

PRDM1 P=2.61 x 10-7

Log H

paII

/MspI

4

2

0

-2

PRDM1 P=0.0004

Gene

expre

ssio

n v

alu

es

for

PR

DM

1

4

4.5

5

XBP1 P=1.63 x 10-7

Log H

paII

/MspI

4

2

0

-2

3

1

-1

Gene

expre

ssio

n v

alu

es

for

XB

P1

XBP1 P=0.0014

6

8

7

9

10

ERN1 P=0.0001

Log H

paII

/MspI

1

0

-1

-2

Gene

expre

ssio

n v

alu

es f

or

ER

N1

4.6

5

5.4

5.8

ERN1 P=7.7 x 10-7

HSPA5 P=1.8 x 10-4

Log H

paII

/MspI

4

2

0

-2

Gene

expre

ssio

n v

alu

es

for

HS

PA

5

5.5

6.0

6.5

7.0

HSPA5 P=0.018

5.0

Pre-B cells/Pre-B ALL B cell lymphoma Pre-B cells/Pre-B ALL Mature B cells/B cell lymphoma

h

i

j

k

Xbp1s

Fig. S3 legend: The gene expression profiling of UPR-related genes (ERN1 (IRE1), HSPA5 (BiP),PRDM1 (Blimp1) and XBP1 microarray data (GEO accession numbers GSE45460, GSE19599,GSM488979, GSE13411, GSE12453 is shown (a). Xbp1s mRNA levels were measured by qRT-PCRusing specific primers in Pro-B and Pre-BI cells using sorted Lin- Sca-1+ c-Kit+ (LSK) multilineageprogenitors, and bone marrow B cell precursor populations as well as in mature B cells from spleens ofC57BL/6 mice (n=3; b). Xbp1s mRNA levels were measured by qRT-PCR after reconstitution of BCR-ABL1 Ighm-/- pre-B ALL cells with an Ig µ-heavy chain expression vector (µHC) and empty vectorcontrol (EV; n=3; c). The methylation of the promoter regions of ERN1, HSPA5, PRDM1 and XBP1 genesin ALL subsets compared to mature B cell Non-Hodgkin lymphomas and pre-B cells was assessed byHELP assay (GEO number GSE34937) and the p-values were calculated from Wilcoxon test (d-g). themRNA levels of ERN1, HSPA5, PRDM1 and XBP1 genes from gene expression microarray data indifferent ALL subsets in comparison to B-cell subsets, Hodgkin and B-NHL and the p-value wascalculated by Wilcoxon test (St Jude http://www.stjuderesearch.org/data/ALL3; h-k)

Fig. S4: Sorting strategies of different human and murine B-cell fractions

CD43

B220

CD24B

P-1

NK1.1 + GR1

FS

C

IgM

IgD

Lineage

FS

C

Sca-1

c-k

it

Fig. S4 legend: Multilineage progenitors, and bone marrow Bcell precursor populations as well as mature B cells from spleensof C57BL/6 mice and healthy human donors were sortedaccording to Hardy fractions A-F using multiple surface markersincluding B220, CD43, BP-1, CD24, NK1.1/Gr1, IgD, IgM,Lineage, c-kit and Sca-1 with a detailed gating strategy as shownin the figure (n=5.a). The different fractions of MACS-enrichedCD19+ pre-B cells (original mononuclear cells: MNC; flow-through and positively selected CD19+ cells: CD19+ MACS)were stained with the surface markers CD22 and CD19 withthree patient-derived pre-B ALL cases (ALL-X2; ALL-Q5; ALL-O2) as a positive control (n=3; b).

CD19

CD

22

MNC Patient-derived pre-B ALL

Flow-through

CD19+ MACS

a

b

0.35

Prd

m1

mR

NA

leve

ls [%

Hprt

]

0.3

0.25

0.2

0.15

0.1

0.05

0

P=0.007 P=0.001

2.5

Xb

p1s

mR

NA

leve

ls [%

HP

RT

]

2

1.5

1

0.5

0

P=0.0005 P=0.002

0.35

0.3

0.25

0.2

0.15

0

0.1

0.05

Prdm1fl/fl ALL

Cre EV Cre

2 days 5 days

EV Cre EV Cre

2 days 5 days

EV

Prdm1fl/fl ALL

Multiple Myeloma

CD19

CD

138

Patient-derivedPre-B ALL

7 AAD

An

nexi

nV

18

19

EV

Cre

Day 1 Day 2 Day 5

14

20

18

25

Prdm1fl/fl ALL

51 XBP1s

b-actin

BM CD19+ pre-B cells Patient –derived Pre-B ALL

PRDM1

Fig. S5: Prdm1 (Blimp1) is required for proliferation and survival of pre-B ALL cells

a

c d e

f

7 AAD

Brd

US 39

S 24

G2/M 3.5

G0/1 57

G0/1 71.5

G2/M 4.3

100

80

60

40

20

0

P=0.0001

S phaseDay 5

Prdm1fl/fl ALL

EV

Cre

EV Cre

G2/MS phase

G0/1

g

200300

WT WT EV Cre EV Cre H20bp

Prdm1fl/fl ALLb

Prdm1flox

Prdm1

24 48 hours

Cre EV Cre

Day 1 Day 2

EV Cre

Day 5

Nu

mber

of

ap

opto

tic c

ells

[%

]

P=0.25 P=0.0004 P=0.0002

0

5

10

15

20

25

EV

Figure S5 legend: The protein levels of PRDM1 and spliced XBP1 (XBP1s) were measured by Westernblot analysis in MACS-enriched normal human CD19+ pre-B cells compared to patient-derived pre-BALL cells and analyzed by densitometry (ALL-Q5, ALL-LAX9, ALL-X2, ALL-BLQ11, ALL-O2) usingβ-actin as loading control (n=3; a). Genomic deletion was assessed after 24 hours treatment with 4-OHTusing specific primers to amplify the floxed allele in Prdm1fl/fl ALL cells transduced with empty vectorcontrol (EV) and 4-OHT-inducible retroviral Cre (Cre) using a wildtype (WT) and H2O control (n=2)(b). Prdm1 mRNA levels were assessed in Prdm1fl/fl ALL cells with EV and Cre at 2 and 5 days aftertreatment with 4-OHT (n=3) (c). Likewise, the Xbp1s mRNA levels were measured in Prdm1fl/f ALLcells with EV and Cre by qRT-PCR after 2 and 5 days post 4-OHT treatment (n=3; d). A representativeFACS blot is shown for a patient-derived Pre-B ALL sample (ALL-ICN1) and a multiple myeloma cellline (JJN3) stained for the surface markers CD138 and CD19 (e). Apoptosis was assessed by Annexin-Vand 7AAD FACS staining in Prdm1fl/f ALL cells with EV and Cre after 1, 2 and 5 days treatment with 4-OHT (n=3; f). Prdm1fl/f ALL cells with EV and Cre were stained with BrdU and 7AAD and thepercentages of cells in the G0/1, S and G2/M cell cycles are indicated after 5 days of 4-OHT treatment andshown in bar graph (n=3, p=0.0001 for the S-Phase; g).

XB

P1s

Ra

tio D

ensi

tom

etr

y [%

]

P=0.03

0

1

2

3

4

PR

DM

1 R

atio

Densi

tom

etr

y [%

]

P=0.06

0

1

2

3

4

Fig. S6: Regulation of XBP1 in Ph+ ALL

Input

BACH2

0

BCL6

16

0

16

0

XBP1

16

0

2 kb

BACH2, BCL6 ChIP-seq; XBP1 promoter

Log r

atio

BC

L6 C

hIP

/ In

put

XBP1

Imatinib

0 mmol/l

HPRT HPRT

Human Ph+ ALL, BCL6 ChIP-on-chip

-1 -0.5 -0 kb -1 -0.5 -0 kb -1 -0.5 -0 kb from TSS

-1 -0.5 -0 kb -1 -0.5 -0 kb -1 -0.5 -0 kb from TSS

0 mmol/l

2

1

0

2

1

0

-1 -0.5 -0 kb -1 -0.5 -0 kb -1 -0.5 -0 kb from TSS

-1 -0.5 -0 kb -1 -0.5 -0 kb -1 -0.5 -0 kb from TSS

Imatinib

10 mmol/l

10 mmol/l

XBP1

2

1

0

2

1

0

0 10 0 10 0 10 0 10 IM [µM]

b-actin

BACH2

Human Ph+ ALL Mouse BCR-ABL1pre-B ALL

c

d

Fig S6 legends: Protein levels of phosphorylated and total ERK were measured by Western blotanalysis in human Ph+ ALL cells (PDX2, LAX7R) treated with or without the MEK-inhibitorPD325901 for 5 hours (50 nM PD325901) using β-actin as loading control (a). Protein levels ofphosphorylated and total S6 were measured by Western blot analysis in human Ph+ ALL cells (PDX2,LAX7R) treated with or without the AKT-inhibitor AZD5363 for 5 hours (10 μM AZD5363) using β-actin as loading control (b). Protein levels of BACH2 were measured by Western blot analysis inhuman Ph+ ALL cells (ICN1, BV-173, TOM1) and mouse BCR-ABL1 pre-B ALL cells treated with orwithout the TKI imatinib for 16 hours (human: 10 µM IM; mouse 2 µM IM) using β-actin as loadingcontrol (c). ChIP-seq analysis was performed using antibodies against BACH2 and BCL6 (GEOnumber GSE44420). Peaks with significant enrichment of ChIP-seq relative to input were identified byChIP-seeqer (http://physiology.med.cornell.edu/faculty/elemento/lab/chipseq.shtml) and shown as boldunderlines. The gene model in UCSC genome browser view (hg18) and the arrow for transcription startsite were shown (d). BCL6 ChIP-on-chip analysis was performed on human Ph+ALL cell lines (Tom1,Nalm1, BV-173) treated with and without imatinib (10 µmol-l) for 24 h as previously described (26)(two representatives shown; GEO number GSE24404) (e).

Human Ph+ ALL

0 50 0 50 PD325901 [nM]

β-actin

ERKpT202/Y204

ERK

a

Human Ph+ ALL

0 10 0 10 AZD5363 [μM]

β-actin

S6 pS235/36

S6

b

e

Fig. S7: Verification of Xbp1-deletion in pre-B cells and BCR-ABL1 ALL

+/+ +/+ EV Cre EV Cre E V Cre EV Cre EV Cre EV Cre H20(bp)

400500

Xbp1fl/fl BCR-ABL1 ALL cells

a

EV Cre0

0.4

0.8

1.2

1.6

2

P=0.0049

Xbp1 m

RN

A le

vels

[%

HP

RT

]

Xbp1 m

RN

A le

vels

[%

HP

RT

]

0

1

2

3

4

5

6

P=0.007

EV Cre

b cXbp1fl/fl pre-B cells Xbp1fl/fl BCR-ABL1 ALL cells

Fig. S7 legend: A representative schematic of the designed primers amplifying the loxP sites at exon 2

of the Xbp1 gene is shown in the top panel. In the lower panel, genomic deletion by PCR was assessed

in Xbp1fl/fl ALL cells with EV or Cre after 1 day administration of 4-OHT using a wildtype (WT) and

H2O control (n=6; a). Xbp1 mRNA levels were measured in IL7-dependent Xbp1fl/fl pre-B cells with

EV or Cre after 24 hours of 4-OHT treatment by qRT-PCR (n=3; b). Xbp1 mRNA levels were

measured in Xbp1fl/fl BCR-ABL1 ALL cells with EV and Cre after 24 hours of 4-OHT treatment by

qRT-PCR (n=3; c). Xbp1fl/fl BCR-ABL1 ALL cells were stained for the surface markers CD19, B220, c-

Kit, Sca-1 and CD13 (n=3; d). Apoptosis was quantified by Annexin-V and 7AAD staining in Xbp1fl/fl

NRASG12D ALL cells with EV and Cre after 1, 4 and 7 days treatment with 4-OHT (n=3; e) Apoptosis

was quantified by Annexin-V and 7AAD FACS staining in IL7-dependent Xbp1fl/fl pre-B cells with

(Cre) or without (EV) deletion of Xbp1 after 1, 2 and 5 days treatment with 4-OHT (n=3; f).

B220

CD

19

Sca1

c-K

it

CD13

Sca

-1

Xbp1fl/fl BCR-ABL1 ALL cellsd

Day 1 Day 4

EV Cre

Day 7

Nu

mbe

r of apo

pto

tic c

ells

[%

]

0

5

10

15

20

25

EV CreEV Cre

P= 0.0002 0.001 3.5 X 10-5

30

eXbp1fl/fl NRASG12D ALL

Day 1 Day 4

EV Cre

Day 7

Nu

mbe

r of apo

pto

tic c

ells

[%

]

0

20

40

60

EV CreEV Cre

P= 0.0008 0.06 1.57 X 10-5

80

Xbp1fl/fl Pre-B cells

f

Fig. S8: Pharmacological inhibition of ERN1 endoribonuclease activity causes cell death in patient-derived pre-B ALL cells

Dose

Eff

ect

0 20 40 60 STF-083010 concentrations [μM]

0

20

40

60

80

100Day 0 Day 4

Control

TN 10 µg

ERN1i 5 µM

ERN1i 10 µM

ERN1i 15 µM

ERN1i 30 µM

ERN1i 60 µM

FSC

DA

PI

da Patient-derived Pre-B ALL cells Patient-derived Pre-B ALL cells

Fig. S8 legend: The dose effect was calculated as measuredfor mRNA levels of XBP1s by qRT-PCR with the CalcuSynsoftware in patient-derived pre-B ALL cells (ALL-Q5) after24 hours treatment with either the ER-stress inducerTunicamycin (TN; 10 µg) in combination with STF-083010(0-60 µM) (ERN1i) for 24 hours (n=3; a). Patient-derived pre-B ALL cases (ALL-Q5 and ICN1) were treated with STF-083010 or A106 (0-60 µM) and the relative viability wasassessed by CCK-8 assay (n=3; b). Apoptosis was assessed byAnnexin-V and 7AAD FACS staining in two patient-derivedpre-B ALL cases (ALL-Q5 and ALL-O2) after 3 daystreatment with either STF-083010 and A106 at twoconcentrations (30 and 60 µM) or DMSO control (0 µM; n=3;c). Patient-derived pre-B ALL (ALL-Q5) were treated withA106 (0-60µM) and TN (10µg) and viability was assessed byDAPI staining up to four days. The representative FACS blotsare shown at day 0 and 4 (n=3; d).

0 10 20 30 40 50 60 ERN1 inhibitor [mM]

80

60

40

20

0

Su

rviv

al o

f ce

lls [%

] Pre-B ALL 1 + STF-083010Pre-B ALL 1+ A106Pre-B ALL 2 + STF-083010Pre-B ALL 2 + A106

b

7 AAD

Annexi

nV

0STF-08301030 mM

A10630 mM

6 69 67

18 71 61

Pre

-B A

LL

Pre

-BA

LLc

Nu

mb

er

of apo

pto

tic c

ells

[%

]

0

20

40

60

80

Ctr STF-083010

A106 Ctr STF-083010

A106

3.5 X 10-5 5.3 X 10-5

Fig. S9: Pharmacological inhibition of ERN1 endoribonuclease activity causes cell death in patient-derived pre-B ALL cells

17 66 94

B-NHL,Mantle cell lymphoma

22 38 66

7 31 87

12 86 94

15 34 65

11 76 96

8 7 9

7 AAD

Annexi

nV

6 25 81

ERN1i [μM] 0 30 60

B-NHL,DLBCL

Multiple Myeloma

Pre-B ALL

Pre-B ALL

Pre-B ALL

Ph+ ALL

Ph+ ALL

a

Fig. S9 legend: Apoptosis was assessed by Annexin-V and 7AAD FACS staining in a panel of humancell lines including B-NHL (Toledo, Jeko), multiple myeloma (JJN3) and patient-derived Pre-B and Ph+

ALL cases (BLQ1, ICN1, ALL-O2, ALL-L3, ALL-Q5) after 3 days treatment with STF-083010 at twoconcentrations (30 and 60 µM) or DMSO control (n=3; a)). Patient-derived pre-B ALL cells (ALL-Q5)were treated with either TN (10µg) or STF-083010 (0-60 µM) and stained with ER-Tracker green after24 hours and 48 hours (n=3; b).

TN

DMSO

A106 5 µM

A106 10 µM

A106 15 µM

A106 30 µM

A106 60 µM

48 h24 h

Patient-derived Pre-B ALLb

Fig. S10: High XBP1 mRNA levels at the time of diagnosis predict poor clinical outcomein patients with pre-B ALL

COG P9906 n=207 pre-B ALL

XBP1Low

XBP1High

P=0.005

Time after diagnosis [years]

100

80

60

40

20

0

0 1 2 3 4 5 6

Rela

pse

-fre

e s

urv

iva

l [%

]

XBP1Low BACH2High

XBP1High BACH2Low

COG P9906 n=207 pre-B-ALL

P=0.00064

Time after diagnosis [years]

100

80

60

40

20

0

0 1 2 3 4 5 6

XBP1 probe set COG: Affy plus2, 200670_at: 11 probes

ECOG, XBP1, NimbleGen: HSAP0406S00034989

ER

N1 L

og

2e

xpre

ssio

n v

alu

es

CR+ CR-

n=22 n=14

n=53 adult pre-B ALL

ECOG E2993, ERN1 mRNA levels

P=0.04 P=0.01

n=36 adult Ph+ ALL

1.5

1

0.5

0

-0.5

-1

1.5

1

0.5

0

-0.5

-1

-1.5

CR+ CR-

n=38 n=15

ER

N1 L

og

2e

xpre

ssio

n v

alu

es

Rela

pse

-fre

e s

urv

iva

l [%

]

b

a

c d

e f Fig. S10 legend: The position ofthe different used XBP1 probesetin this study is shown at the3’UTR for the Affy plus2 (11probesets) and NimbleGen (1probeset; a, b). In an analysis,pre-B ALL (COG clinical trialP9906, n=207) patients weresegregated into two groups basedon high or low mRNA levels inrespect to the median mRNAvalue of the XBP1 probeset (c).

In a multivariate analysis, pre-B ALL (COG clinical trial P9906, n=207) patients were segregatedinto two groups based on high or low mRNA levels of XBP1 or BACH2, in respect to the medianexpression value of the XBP1 probeset. (d). ERN1 expression was measured in patients with either apositive (CR+) or negative complete response (CR-) in both adult Ph+ ALL (n=36) and adult pre-BALL patients (n=53) (e, f).

1

Table S1: Overview of patient-derived samples of Ph+ ALL and cell lines __________________________________________________________________________ Case Karyotype Clinical course Gender/Age ___________________________________________________________________________________________ Primary cases, Pre-B ALL ALL-X2 uncharacterized Unknown m/38

LAX9 t(9;22)(q34;q11) At diganosis m

del(12)(p12;p13)

ALL-O2 NRASG12D At diagnosis m/7

BLQ1 FISH der(9), der(22) Unknown

ALL-Q5 KRASG12V Unknown f

BLQ11 FISH der(9), der(22) Relapse (Imatinib) m

ALL-L3 KRASG12V At diagnosis

ICN1 der(9)(q10)t(9;22)(q34;q11) At diagnosis f

LAX7R uncharacterized Relapse

PDX2 BCR-ABL1 p190 At diagnosis f/52 Primary cases, Multiple myeloma MM1 n.d. Newly diagnosed m/44

MM2 n.d. Relapsed m/66

MM3 n.d. Persistent /Relapsed m/66

MM4 n.d. Persistent /Relapsed m/68

MM5 n.d. Newly diagnosed f/36

MM6 n.d. Persistent, at diagnosis m/65 ___________________________________________________________________________________________

Cell lines BV173 add(1)(q42), add(8)(p23), m/45

der(22)t(9;22)(q34;q11)

Nalm1 der(7)t(7;9;15)(q10;?;q15), f/3

t(9;9)(p24;q33-q34) t(9;22)

(q34;q11), der(15)t(7;9;15)

SUP-B15 t(1;1)(p11;q31), add(3)(q27), Relapse m/9

add(10)(q25), t(9;22)(q34;q11)

TOM1 del(7)(p14),der(9)del(9)(q13q34) Refractory f/54

der(22)t(9;22)(q34;q11)

JJN-3 hypotriploid karyotype At diagnosis f/57

with 9% polyploidy

JeKo-1 ATM amplification, f/78

MYC amplification

P16INK4A deletion

SYK amplification

P53 deletion/mutation

Toledo Complex karyotype f ___________________________________________________________________________________________ Notes: All primary samples are bone marrow biopsies, blast content >80%

2

Table S2: Overview of mouse strains used in this study

_________________________________________________________________________________ Mouse strain Source Purpose _________________________________________________________________________________ Bach2-/- Kazuhiko Igarashi, Tohoku University Analysis of Xbp1 as Bach2 target gene

BCL6 -/- B. Hilda Ye, Albert Einstein College Analysis of Xbp1 as BCL6 target gene aXbp1fl/fl Laurie H. Glimcher, Cornell Medical School Genetic loss-of-function experiments bStat5fl/fl Lothar Hennighausen, NIDDK Analysis of Xbp1 as STAT5 target gene Prdm1fl/fl Jackson Laboratories Genetic loss-of-function experiments Hspa5 fl/fl Amy S Lee, USC Genetic loss-of-function experiments

NOD/SCID Jackson Laboratories Xenograft recipient mice ___________________________________________________________________________________________ Notes: a Lee, A.-H., Scapa, E. F., Cohen, D. E. & Glimcher, L. H. Regulation of hepatic lipogenesis by the transcription

factor XBP1. Science 320, 1492–1496 (2008). b Liu, X. et al. Stat5a is mandatory for adult mammary gland development and lactogenesis. Genes Dev. 11, 179–

186 (1997).

3

Table S3: Sequences of oligonucleotide primers used Quantitative RT-PCR

Human

COX6B_F 5’-AACTACAAGACCGCCCCTTT-3’

COX6B_R 5’-GCAGCCAGTTCAGATCTTCC-3’

XBP1s_F 5’- TGCTGAGTCCGCAGCAGGTG -3’

XBP1s_R 5’- GCTGGCAGGCTCTGGGGAAG -3’

ERDJ4_F 5’- GCTACTCCCCAGTCAATTTTCA -3’

ERDJ4_R 5’- CCGATTTTGGCACACCTAAGAT -3’

p58IPK_F 5’- TGTGTTTGGGATGCAGAACTAC -3’

p58IPK_R 5’- TCTTCAACTTTGACGCAGCTT -3’

SEC61A1_F 5’- TGTCATCTCCCAAATGCTCTCA -3’

SEC61A1_R 5’- ACAGGTAATAGCAAAGGCCAC -3’

Mouse

Hprt_F 5’-GGGGGCTATAAGTTCTTTGC-3’

Hprt_R 5’-TCCAACACTTCGAGAGGTCC-3’

Xbp1 (ex2)_F 5’- CCTGAGCCCGGAGGAGAA -3’

Xbp1 (ex2)_R 5’- CTCGAGCAGTCTGCGCTG -3’

ERdj4_F 5’- GGATGGTTCTAGTAGACAAAGG-3’

ERdj4_R 5’- CTTCGTTGAGTGACAGTCCTGC-3’

p58IPK_F 5’- GCATCTTGAATTGGGGAAGA-3’

p58IPK_R 5’- CAAGCTTCCCTTGTTTGAGC-3’

Prdm1_F 5’- GAAATGGAAAGATCTATTCCAGAG-3’

Prdm1_R 5’- GGCATTCTTGGGAACTGTGT-3’

Cox6a2_F 5’- CGGTTATGAGCACCCTTGAT-3’

Cox6a2_R 5’- CTGTTCCCAAAGAGCCAGAG-3’

Ccng2_F 5’-TGATCCGCATCAGTCAGTGT-3’

Ccng2_R 5’- ACAATCGCGTGGTACAAGTG-3’

CD22_F 5’-GAGCAAGCTCACCTTCCAAC-3’

CD22_R 5’-CACCTCTGTGGGATTGACCT-3’

Ccr7_F 5’- GCCTTGATCACCATCCAAGT-3’

Ccr7_R 5’- GATCACCTTGATGGCCTTGT-3’

Cy5r1_F 5’- ATCCAACCCAGTGCTTTCTG-3’

Cy5rl_R 5’- CAAAGCCCTTGCTGTAGGTC-3’

Genomic PCR

Mouse

Xbpt_F 5’-TTTGGCTTGGGGAGGGACA-3’

Xbpt_R 5’-AGCAGTCTGCGCTGCTACTCT-3’

Prdm1_F 5’-CAATGCTTGTCTAGTGTC 3’

Prdm1_R 5’-AGTAGTTGAATGGGAGC-3’

Hspa5_F 5’-GATTTGAACTCAGGACCTTCGGAAGAGCAG-3’

Hspa5_R 5’-TTGTTAGGGGTCGTTCACCTAGA-3’

4

Table S3: Sequences of oligonucleotide primers used –continued

Single locus ChIP PCR

Human

BACH2_F CCTACCTGGCAAAAACAAAAAC

BACH2_R TCTTTTTGAGCAGTGGCATAGA

BCL6_F 5’-CCGAGAATTGAGCTCTGTTGA-3’

BCL6_R 5’- GGCAGCAACAGCAATAATCA-3

XBP1_F AATGTTTTGGGAGAAGTCTGGA

XBP1_R TAAAACGTAACAGGTGGCATGA

GAPDH_F ACGTAGCTCAGGCCTCAAGA

GADPH_R GCTGCGGGCTCAATTTATAG

5

Table S4: Antibodies used for Western blotting and flow cytometry

Western blot

Antigen Clone ID Source

ACTB Polyclonal (ab8227) Abcam

XBP1 Polyclonal (ab37152) Abcam

ARF- Mo Polyclonal (ab80) Abcam

P21 C-19 Santa Cruz Biotechnology

P27 F-8 Santa Cruz Biotechnology

p38 5F11 Cell Signaling Technology

Tp53 IC12 Cell Signaling Technology

p-P38(Thr180/Tyr182) Polyclonal (9211) Cell Signaling Technology

JNK Polyclonal (9252) Cell Signaling Technology

p-JNK(Thr183/Tyr185) Polyclonal (9251) Cell Signaling Technology

CHOP Polyclonal (5554) Cell Signaling Technology

Prdm1 Monoclonal (9115) Cell Signaling Technology

Cre Polyclonal (ab40011) Abcam

pERK(Thr202/Tyr) Monoclonal (2211) Cell Signaling Technology

ERK Polyclonal (ab40011) Abcam

pS6(Ser235/236) Polyclonal (2211) Cell Signaling Technology

S6 Monoclonal (2217) Cell Signaling Technology

Flow cytometry

Surface Antigen Source

CD19 BD Biosciences

Sca-1/Ly6a BD Biosciences

CD117/cKIT BD Biosciences

CD138/SDC1 BD Biosciences

B220/CD45R BD Biosciences

CD43 Biolegend

BP-1 Biolegend

NK1.1 and GR1 Biolegend

IgD Biolegend

IgM Biolegend

Lin Biolegend

CD24 BD Biosciences

Annexin-V BD Biosciences

7AAD Invitrogen

TruStain FxX block BD Biosciences

Lin (CD3/33/56) BD Biosciences

CD34 Biolegend

CD10 BD Biosciences

CD19 Invitrogen

CD22 BD Biosciences

Human Fc block Miltenyi

BrdU Invitrogen

6

Table S5: Retroviral and lentiviral vectors used

Constitutive expression Inducible activation MSCV BCR-ABL1-IRES-Neo MSCV-Cre-ERT2

MSCV-GFP-IRES MSCV-ERT2________________

MSCV-GFP-IRES-CRE MSCV-Cre-ERT2-GFP

MI µHC-CD8 MSCV-ERT2-GFP___________

MI CD8 PAX5-ERT2-GFP

pCCL-c-MNDU3-Luc-SV40-Neo pcl6-ERT2-GFP_____________