Biological implications of somatic DDX41 p.R525H mutation in acute myeloid leukemia

Accepted Manuscript Biological implications of somatic DDX41 p.R525H mutation in acute myeloid leukemia Moe Kadono, Akinori Kanai, Akiko Nagamachi, Satoru Shinriki, Jin Kawata, Koji Iwato, Taiichi Kyo, Kumi Oshima, Akihiko Yokoyama, Takeshi Kawamura, Reina Nagase, Daichi Inoue, Toshio Kitamura, Toshiya Inaba, Tatsuo Ichinohe, Hirotaka Matsui PII:

S0301-472X(16)30125-4

DOI:

10.1016/j.exphem.2016.04.017

Reference:

EXPHEM 3399

To appear in:

Experimental Hematology

Received Date: 13 April 2016 Revised Date:

28 April 2016

Accepted Date: 29 April 2016

Please cite this article as: Kadono M, Kanai A, Nagamachi A, Shinriki S, Kawata J, Iwato K, Kyo T, Oshima K, Yokoyama A, Kawamura T, Nagase R, Inoue D, Kitamura T, Inaba T, Ichinohe T, Matsui H, Biological implications of somatic DDX41 p.R525H mutation in acute myeloid leukemia, Experimental Hematology (2016), doi: 10.1016/j.exphem.2016.04.017. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Graphical abstract DDX41 p.R525H

Ribosomal proteins

Defect in pre-rRNA processing

Low mature rRNA

RPLs

28S

RPSs

5.8S 5S

Impaired ribosomal biogenesis 60S 40S

Free ribosomal proteins RPL5 RPL11

SC

RB

RB activation Cell cycle inhibition

RI PT

18S

RPL5

RPL11

M AN U

MDM2

AC C

EP

TE D

MDM2

MDM2

ACCEPTED MANUSCRIPT

1

Article Title

2

Biological implications of somatic DDX41 p.R525H mutation in acute myeloid

3

leukemia

RI PT

4 5

Authors

6 7 8 9

Moe Kadono1,2, Akinori Kanai1, Akiko Nagamachi1, Satoru Shinriki3,4, Jin Kawata4,

20 21 22 23 24 25 26 27 28 29 30 31 32

SC

M AN U

1. Department of Molecular Oncology and Leukemia Program Project, Research Institute for Radiation Biology and Medicine, Hiroshima University, Hiroshima, Japan.

2. Department of Hematology and Oncology, Research Institute for Radiation Biology and Medicine, Hiroshima University, Hiroshima, Japan. 3. Department of Molecular Laboratory Medicine, Graduate School of Medical

TE D

16 17 18 19

and Hirotaka Matsui1,3,4*

Sciences, Kumamoto University, Kumamoto, Japan. 4. Central Clinical Laboratory, Kumamoto University Hospital, Kumamoto, Japan. 5. Department of Hematology, Hiroshima Red Cross Hospital & Atomic-bomb Survivors Hospital, Hiroshima, Japan.

EP

13 14 15

Reina Nagase8, Daichi Inoue8, Toshio Kitamura8, Toshiya Inaba1, Tatsuo Ichinohe2*

6. Laboratory for Malignancy Control Research, Medical Innovation Center, Kyoto University Graduate School of Medicine, Kyoto, Japan. 7. Department of Molecular Biology and Medicine, Laboratory for System Biology

AC C

10 11 12

Koji Iwato5, Taiichi Kyo5, Kumi Oshima2, Akihiko Yokoyama6, Takeshi Kawamura7,

and Medicine (LSBM), Research Center for Advanced Science and Technology (RCAST), The University of Tokyo, Tokyo, Japan.

8. Division of Cellular Therapy, The Institute of Medical Science, The University of Tokyo, Tokyo, Japan.

* Corresponding authors *Addresses correspondence to: Hirotaka Matsui E-mail: [email protected]

Tel: +81-96-373-5890, FAX: +81-96-373-5283

1

ACCEPTED MANUSCRIPT

1 2

Department of Molecular Laboratory Medicine, Graduate School of Medical

3 4 5 6

1-1-1 Honjo, Chuo-ku, Kumamoto, Japan 860-8556

Sciences, Kumamoto University

Tatsuo Ichinohe Tel: +81-82-257-5861, FAX: +81-82-256-7108

RI PT

E-mail: [email protected]

7 8 9 10

Department of Hematology and Oncology, Research Institute for Radiation Biology

11 12 13 14

Word count for the text: 3,315 words

15

Highlights

and Medicine, Hiroshima University

SC

1-2-3 Kasumi, Minami-ku, Hiroshima, Japan 784-8553

Figures: 4, References: 28

M AN U

Scientific category: Malignant hematopoiesis Key words: DDX41 p.R525H, pre-rRNA processing, Acute myeloid leukemia

16 17

DDX41 p.R525H mutation causes a defect in pre-rRNA processing.

18 19

These mechanisms account for the development of AML exhibiting cytopenias.

22 23 24 25 26

TE D EP

21

AC C

20

DDX41 p.R525H inhibits cell cycle progression through MDM2-RB-E2F axis.

27 28 29

2

ACCEPTED MANUSCRIPT

1

Abstract The DDX41 gene, encoding a DEAD-box type ATP-dependent RNA helicase,

3

is rarely but reproducibly mutated in myeloid diseases. The acquired mutation in

4

DDX41 is highly concentrated at c.G1574A (p.R525H) in the conserved motif VI

5

located at the C-terminus of the helicase core domain where ATP interacts and is

6

hydrolyzed. Therefore, it is likely that the p.R525H mutation perturbs the ATPase

7

activity in a dominant-negative manner. In this study, we screened for the DDX41

8

mutation of CD34-positive tumor cells based on mRNA-sequencing and identified

9

the p.R525H mutation in three cases among 23 patients. Intriguingly, these

10

patients commonly exhibited acute myeloid leukemia (AML) with peripheral blood

11

cytopenias and low blast counts, suggesting that the mutation inhibits growth and

12

differentiation of hematopoietic cells. Data from cord blood cells and leukemia cell

13

lines suggest a role for DDX41 in the pre-ribosomal RNA processing, in which the

14

expression of p.R525H mutant causes a certain ribosomopathy phenotype in

15

hematopoietic cells by suppressing MDM2-mediated RB degradation, thus

16

triggering the inhibition of E2F activity. This study uncovered a pathogenic role of

17

p.R525H DDX41 in the slow growth rate of tumor cells. Age-dependent epigenetic

18

alterations or other somatic changes might collaborate with the mutation to develop

AC C

EP

TE D

M AN U

SC

RI PT

2

1

ACCEPTED MANUSCRIPT

1

AML.

2

RI PT

3 4

SC

5 6

M AN U

7 8

12 13 14 15

EP

11

AC C

10

TE D

9

16 17 18 2

ACCEPTED MANUSCRIPT

1

Introduction Current comprehensive sequencing approaches led to the identification of

3

rare but reproducible somatic gene mutations in myeloid malignancies. Among

4

them, there is a somatic mutation in the DDX41 gene encoding a DEAD-box type

5

ATP-dependent RNA helicase. The somatic mutation in DDX41 is highly

6

concentrated at c.G1574A (p.R525H) in the conserved motif VI located at the

7

C-terminus of the RecA-like helicase core domain where ATP interacts and is

8

hydrolyzed. Therefore, it is likely that the p.R525H mutation in the DDX41 protein

9

perturbs the ATPase activity in a dominant-negative manner. In addition, recently

10

germ line mutations in DDX41 were isolated in a subset of familial

11

AML/myslodysplastic syndrome (MDS) pedigrees.[1,2] Since roles of these somatic

12

and germ line mutations in the pathogenesis of myeloid diseases are not completely

13

understood, researchers who first described these mutations advocate a role for

14

DDX41 in mRNA splicing by interacting with the U2 and U5 complexes.[1]

15

However, in contrary to canonical spliceosomal mutations, the DDX41 mutation

16

develops not only MDS, but also primary AML.

AC C

EP

TE D

M AN U

SC

RI PT

2

17

In this study, we propose a role for DDX41 as a precursor ribosome RNA

18

(pre-rRNA) processing factor, in which the p.R525H mutation affects ribosome

3

ACCEPTED MANUSCRIPT

biogenesis. We found that ribosome biogenesis was widely affected when cord

2

blood-derived CD34-positive cells were transfected with DDX41 p.R525H, thus

3

compromising cell cycle progression through impaired E2F function. Since

4

molecular mechanisms for the development of AML with cytopenias have not been

5

clearly elucidated, in this report we propose that the DDX41 p.R525H mutation in

6

hematopoietic stem/progenitor cells are involved in the pathogenesis of a certain

7

subset of such AML.

8

10

Materials and Methods

Patient samples and cell fractionation fractionation

TE D

9

M AN U

SC

RI PT

1

Bone marrow aspirates or peripheral blood specimens were collected from

12

23 patients with AML who participated in the study, according to the protocol

13

approved by the ethics committee on the human genome research at Hiroshima

14

University and Kumamoto University. The protocol was based on an opt-out or on

15

written informed consent obtained by the patients. The protocol included the use of

16

pooled samples at initial diagnosis. Total RNA was extracted from CD34-positive

17

cell fractions isolated by magnetic activated cell sorting. Where possible,

18

CD3-positive and CD34-negative/CD3-negative fractions were isolated.

AC C

EP

11

19 20

mRNA sequencing

4

ACCEPTED MANUSCRIPT

In this study, gene mutations were screened upon mRNA sequencing. The

2

libraries for mRNA sequencing were prepared according to the SureSelect library

3

preparation kit (Agilent Technology, Santa Clara, CA, USA) and subjected to

4

massively parallel sequencing with GAIIX or Hiseq2500 sequencer (Illumina, San

5

Diego, CA, USA) using a single-end 36-bp or 50-bp sequencing length protocol.

6

Sequenced tags were aligned to the human reference genome (build hg19) using

7

ELAND (Illumina), and gene expression was normalized to the amount of reads per

8

kilobase of exon per million mapped (rpkm).

M AN U

SC

RI PT

1

9 Antibodies

TE D

10

The following antibodies were used in this study: anti-DDX41 (ab182007;

12

Abcam, Cambridge, UK), anti-actin (MAB1501; Merck Millipore, Darmstadt,

13

Germany),

14

anti-Myc-tag (#2272; Cell Signaling Technology, Danvers, MA, USA), anti-Nucleolin

15

(M019-3S, MBL, Nagoya, Japan), anti-RPL5 (ab86863; Abcam), anti-RPL11

16

(ab79352; Abcam), anti-RB (#9313; Cell Signaling Technology), anti-phospho-RB

17

(#9301;

18

anti-Multi-Ubiquitin (D058-3; MBL).

EP

11

M2

(F3165;

Sigma-Aldrich,

St-Louis,

MO,

USA),

AC C

anti-FLAG

Cell

Signaling

Technology),

5

anti-MDM2

(ab3110;

Abcam)

and

ACCEPTED MANUSCRIPT

1 2

Vector construction DDX41 and other cDNAs used in this study were obtained by PCR

4

amplification from human cord blood-derived total RNA and cloned into pMYs-IG

5

retrovirus vector.[3] Point mutations and truncations were generated by

6

site-directed mutagenesis approach.

8

M AN U

7

SC

RI PT

3

Cell culture and plasmid transfection

Cord blood-derived CD34-positive cells were purchased from RIKEN Cell

10

Bank (Ibaraki, Japan). The cells were cultured on Tst4/min feeder cells (kindly

11

provided by Dr. H. Kawamoto at Kyoto University) in DMEM supplemented with

12

20% fetal bovine serum (FBS) in the presence of human stem cell factor (SCF),

13

FMS-related tyrosine kinase 3 ligand (FLT3-L), and thrombopoietin (TPO) (100

14

ng/mL each). THP-1 and K562 cells were cultured in RPMI1640 containing 10%

15

FBS. For enforced gene expression, pMYs-IG retroviral vector and PLAT-E or

16

PLAT-F packaging cells were used.

AC C

EP

TE D

9

17 18

Immunofluorescent analysis

6

ACCEPTED MANUSCRIPT

Murine lung fibroblasts grown on coverslips were infected with a retrovirus

2

and were cultured for two days. THP-1 cells were first transduced with retrovirus

3

and were then attached on a slide glass using a cytospin apparatus. The cells were

4

fixed with 4% paraformaldehyde, permeabilized with 0.1% Triton X-100 and blocked

5

with 1% bovine serum albumin, followed by the staining with primary antibodies.

6

After the staining with Cy3- or Alexa488-labelled secondary antibodies and with

7

Hoechst 33342, fluorescent signals were observed, and the images were taken using

8

a confocal laser scanning microscope (LSM 5, Carl Zeiss Microscopy, Jena,

9

Germany).

11

SC

M AN U

TE D

10

RI PT

1

Immunoblot and immunoprecipitation immunoprecipitation analysis Cells (1 × 106) were extracted with NP40 lysis buffer (50mM Tris-HCl pH

13

8.0, 150mM NaCl and 1.0% NP40) containing proteinase and phosphatase inhibitor

14

cocktails (Complete and PhosSTOP; Roche Life Sciences, Indianapolis, IN, USA).

15

For immunoprecipitation analysis, protein G-coated magnet beads (20 µL in each

16

experiment) conjugated with an antibody (4 µg) were mixed with cell extracts

17

overnight at 4°C. The beads were washed five times with NP40 lysis buffer, and

18

then resuspended with 50 µL sample buffer (NP40 lysis buffer containing 1%

AC C

EP

12

7

ACCEPTED MANUSCRIPT

sodium dodecyl sulfate, 0.5% β-mercaptoethanol and 12.5% glycerol), followed by

2

denaturation and elution by heating at 95°C for 2 min. The samples were loaded

3

and electrophoresed on 8.5%, 12.5% or 15% polyacrylamide gels according to the

4

molecular weight of the protein of interest.

RI PT

1

6

SC

5 Northern blotting analysis

Northern blotting analysis was performed based on a non-RI protocol

8

according to the manufacturer’s instructions (DIG Northern Starter Kit; Roche Life

9

Sciences). Briefly, 1.5 µg of total RNA was denatured and fractionated using a

10

denaturing gel (2% formaldehyde/1.2% agarose), followed by transfer to a nylon

11

membrane. The membranes were hybridized with Digoxigenin-labelled RNA probes,

12

and the signals were detected using alkaline phosphatase-labelled anti-Digoxigenin

13

antibody and a chemiluminescent substrate. RNA probes in this assay were

14

prepared by PCR amplification of internal transcribed spacer (ITS) 1 and ITS2

15

regions (for ITS1, forward primer: 5’-acggagcccggagggcgaggcccgc-3’, reverse primer:

16

5’-cgtctccctcccgagttctcggctc-3’;

17

ctaagcgcagacccggcggcgtccg-3’, reverse primer: 5’- acgggaactcggcccgagccggctc-3’),

18

followed by in vitro transcription using DIG-11-UTP for probe labelling. The ITS1

AC C

EP

TE D

M AN U

7

and

for

8

ITS2,

forward

primer:

5’-

ACCEPTED MANUSCRIPT

probe was used for the identification of 47S, 45S, 30S and 21S pre-rRNA, whereas

2

ITS2 probe was for 47S, 45S, 32S and 12S detection (Fig. S1A and S1B). Schematic

3

pre-rRNA processing pathways are shown in Fig. S1B.

4 ATPase assay

SC

5

RI PT

1

The assay reaction (10 µL) contained the following components: 50mM

7

Tris-HCl (pH 7.4), 20mM MgCl2, 50mM KCl, 3.33pM [α-32P]ATP, and 0.2 µg of

8

helicase domain of DDX41 purified by an in vitro translation system (Trans-direct

9

insect cell, Shimadzu Corporation, Kyoto, Japan). Bovine serum albumin (BSA) was

10

used as a negative control. Samples were incubated for 15 or 30 min at 37°C, and

11

the reaction was stopped by adding 1 µL of 0.5M EDTA (pH 8.0). The reaction

12

mixture (1 µL) was then spotted on a cellulose PEI-F plate (Avantor Performance

13

Materials, Center Valley, PA, USA) and developed by ascending chromatography in

14

0.75M LiCl/1M formic acid solution for 1 h. The plate was air dried, and the

15

ATP/ADP signals were detected by autoradiography.

AC C

EP

TE D

M AN U

6

16 17 18

Mice bone marrow transplantation (BMT) C57BL/6 (Ly5.1) mice (Sankyo Labo Service Corporation, Tokyo, Japan) and

9

ACCEPTED MANUSCRIPT

C57BL/6 (Ly5.2) mice (Charles River Laboratories Japan, Kanagawa, Japan) were

2

used for BMT experiments. A total of 100,000 cells (Ly5.1) transduced with a

3

DDX41/GFP-expression vector was transplanted into sublethally irradiated

4

recipient mice (Ly5.2), and peripheral white blood cell (WBC) numbers and GFP

5

signals were measured two to three months later.

SC

RI PT

1

6 Results

8

Identification of DDX41 p.R525H mutation in three AML patients.

M AN U

7

We screened for the DDX41 mutation of CD34-positive tumor cells based on

10

mRNA sequencing and identified heterozygous DDX41 c.G1574A (p.R525H)

11

mutation (chr5: 176,939,370 on build GRCh37/hg19) in three cases among 23

12

patients (Fig. 1A). While we did not perform comprehensive sequencing of paired

13

germline DNA, we confirmed the mutation as somatic by the direct genome

14

sequencing of CD3-positive cells (Fig. 1B). Intriguingly, these patients, in addition

15

to a patient with the same mutation enlisted in The Cancer Genome Atlas,[4]

16

exhibited AML with peripheral and bone marrow cytopenias and low blast counts

17

(Fig. 1A), although precise cellularity of the bone marrow was undetermined

18

because bone marrow biopsy was not performed. In addition, a recent report

AC C

EP

TE D

9

10

ACCEPTED MANUSCRIPT

1

described

2

hypocellularity and low WBC count in a primary AML patient.[2] According to our

3

transcriptome data on patients’ CD34-positive cells, almost half of the sequenced

4

tags at c.G1574 were considered mutated in each patient (18 out of 37, 19 out of 32,

5

and 23 out of 47 reads were called “A”). Therefore, almost all CD34-positive cells in

6

these patients may harbor heterozygous DDX41 p.R525H mutation, though deeper

7

sequencing analysis is required to assess the precise variant allele frequencies. On

8

the other hand, CD34-negative cells carrying the mutation were fewer compared

9

with CD34-positive cells, and CD3-positive cells carrying the mutation were almost

10

absent, as estimated by direct genomic DNA sequencing (Fig. 1B). These data

11

suggest a somatic nature of the mutation in the patient and a block in the

12

differentiation of CD34-positive DDX41 mutant cells into negative cells.

14 15

p.R525H

mutation

associated

with

bone

marrow

EP

TE D

M AN U

SC

RI PT

germline

AC C

13

a

Nuclear localization localization of DDX41 protein We next clarified the localization of DDX41 protein. In this experiment,

16

FLAG-tagged STING, a protein that had been reported to interact with DDX41 at

17

the endoplasmic reticulum,[5,6] was co-expressed with Myc-tagged DDX41 in

18

fibroblasts. Ectopically expressed DDX41 was mostly nuclear regardless of the

11

ACCEPTED MANUSCRIPT

p.R525H mutation and hardly co-localized with FLAG-STING in a steady state (Fig.

2

2A). In the THP-1 leukemia cell line, the endogenous DDX41 protein was also

3

mostly nuclear (Fig. 2B). In the nucleus, DDX41 protein localized both in the

4

nucleoplasm and in the nucleolus (stained with Nucleolin), suggesting that DDX41

5

plays a role in the nucleoplasm as well as in the nucleolus. While previous reports

6

described DDX41 as a cytosolic DNA sensor that recognizes nucleic acids of

7

pathogens,[6] our study did not suggest this function based on the protein

8

localization. However, we identified a 52-kDa DDX41 short isoform, translated from

9

the 2nd methionine, in addition to the full-length 70-kDa DDX41 (Figs. 2C, S2A and

10

S2B). This short form, which lacks putative nuclear localizing signal (NLS) and is

11

detected both in the cytoplasm and in the nucleus (Figs. 2C and S2C), might

12

function as a sensor of pathogenic nucleic acids.

SC

M AN U

TE D

EP AC C

13

RI PT

1

14

Growth inhibition of p.R525H DDX41 expressing cells due to the impaired

15

prepre-rRNA processing

16

To investigate the molecular functions of DDX41 in hematopoietic cells,

17

cord blood-derived human CD34-positive cells were transduced with wild-type (WT)

18

or p.R525H DDX41 using a retroviral system,[3] followed by cultivation in the

12

ACCEPTED MANUSCRIPT

presence of 100 ng/mL SCF, TPO, and FLT3-L. After a 30-day culture, p.R525H cells

2

showed decreased proliferation compared with WT cells (Figure 3A), accompanied

3

by the suppression of mRNAs encoding ribosomal proteins (Fig. S3A). A gene set

4

enrichment analysis (GSEA) [7] showed that gene expression pattern of p.R525H

5

cells negatively correlated with ribosomal gene sets and with myc target genes (Fig.

6

S3B). GSEA further indicated an enrichment of the genes upregulated in the

7

absence of RPS14 in p.R525H cells (Fig. S3B). These data suggest that a certain

8

ribosomopathy[8] may occur in the mutant DDX41 expressing cells.

M AN U

SC

RI PT

1

These findings and a highly conserved DEAD-box type RNA helicase

10

domain of DDX41 led us to speculate that DDX41 protein might be involved in

11

pre-rRNA processing. Actually, a recent paper also supports this speculation.[9] As

12

expected from the position of the somatic mutation (p.R525), which is within the

13

helicase core where the ATP binds and is hydrolyzed (Fig. 2C),[10] the mutant

14

helicase domain displayed a lower ATPase activity (Fig. 3B). In addition, Northern

15

blot analysis probing ITS1 and ITS2 of pre-rRNA showed increased signals of 47S

16

and 41S pre-rRNAs (Fig. 3C); the 30S signal was slightly upregulated, whereas the

17

21S signal was decreased in THP-1 cells expressing p.R525H DDX41. Instead, the

18

32S signal was increased both in WT and p.R525H cells. Furthermore,

AC C

EP

TE D

9

13

ACCEPTED MANUSCRIPT

semi-quantitative RT-PCR experiments showed relatively higher amounts of

2

pre-rRNA containing 5’ external transcribed spacer (ETS) or ITS2 than that

3

containing ITS1 in p.R525H cord blood cells (Fig. 3D). Although the precise phase at

4

which DDX41 takes part in the pre-rRNA processing has not been elucidated, this

5

series of experiments suggests a role for DDX41 in the trimming of 5’ ETS and/or

6

ITS 2.

SC

RI PT

1

8

M AN U

7

Activation of RB pathway occurring in p.R525H DDX41 expressing cells As mentioned before, p.R525H DDX41 cord blood cells were fewer than WT

10

cells after the culture (Fig. 3A), but the induction of apoptosis was unlikely because

11

cell viability was maintained at almost 100% throughout the culture (data not

12

shown). Recent studies on ribosomopathies revealed an activation of the MDM2-p53

13

pathway in the pathogenesis of the diseases.[11,12] It is now widely recognized that

14

RPL5 and RPL11, not incorporated into the 60S ribosome, preferentially bind to

15

MDM2; therefore, inhibition of p53 by MDM2 is compromised, thus causing a

16

stimulation of the p53 pathway.[13,14] We initially assumed that the same defect

17

might be involved in the growth impairment of p.R525H DDX41 cells. Nevertheless,

18

GSEA did not indicate p53 activation in the cells, probably because we cultured cord

AC C

EP

TE D

9

14

ACCEPTED MANUSCRIPT

blood cells with a combination of cytokines (SCF, FLT3-L and TPO), which is

2

expected to expand hematopoietic stem/progenitor cells with less differentiation.

3

Data from transcriptome analysis suggested no erythroid differentiation in this

4

culture condition, in which globin genes were almost not expressed at all (data not

5

shown). Given that ribosomal defect-mediated activation of p53 pathway

6

preferentially occurs in the erythroid lineage in ribosomopathies, our culture

7

condition might not be optimal to observe p53 activation.

M AN U

SC

RI PT

1

On the other hand, GSEA instead revealed a negative enrichment of cell

9

cycle-promoting genes regulated by the RB-E2F axis in p.R525H DDX41 cord blood

10

cells (Fig. 4A). Cell cycle inhibition through the suppression of E2F activity was also

11

detected in patient-derived samples (Fig. 4B). In our study, we found increased

12

RPL5 and RPL11 bound to MDM2 in p.R525H DDX41 cells (Fig. 4C). Since the total

13

amount of these ribosomal proteins was not apparently altered by the enforced

14

expression of p.R525H DDX41, the mutant DDX41 might increase free ribosomal

15

proteins that are not incorporated into the 60S ribosome; these proteins may

16

eventually form a complex with MDM2. Moreover, the RB protein was increased in

17

p.R525H DDX41 cells, in part due to the inhibition of MDM2-mediated

18

poly-ubiquitination (Fig. 4D). The RB protein was active in p.R525H DDX41 cells

AC C

EP

TE D

8

15

ACCEPTED MANUSCRIPT

because it was dephosphorylated, though currently the mechanism of the

2

dephosphorylation is unclear. Furthermore, impaired growth was observed in

3

p.R525H DDX41 THP-1 cells lacking WT p53 (Fig. S4A),[15] thus implying a

4

p53-independent but RB-dependent cell-cycle inhibition by the mutant DDX41.

RI PT

1

6

SC

5 Discussion

In this study, we performed mRNA sequencing on a total of 23 patients and

8

identified the DDX41 p.R525H mutation in three AML patients. The rate of somatic

9

DDX41 mutation was high compared with previous reports because of our limited

10

number of patients or because of the use of CD34-positive cells in our study. In most

11

previous studies, whole genome or exome sequencing was performed using total

12

bone marrow or peripheral blood specimens. Given that tumor cells harboring the

13

DDX41 mutation rarely divide and differentiate, a screening using whole

14

hematopoietic cells might fail to detect mutations of such a rare population, thus

15

underestimating the rate of DDX41 mutation among AML/MDS patients.

AC C

EP

TE D

M AN U

7

16

DDX41 protein is a DEAD-box type ATP-dependent RNA helicase[10],

17

whose helicase domain is highly conserved. DEAD-box RNA helicases have been

18

shown to act in many pathways including pre-mRNA splicing and ribosome

16

ACCEPTED MANUSCRIPT

biogenesis. In addition, recent works propose roles of the RNA helicases, such as

2

RIG-I, in sensing nucleic acids of microbes incorporated into cells.[16,17] Besides, it

3

is noteworthy that some DEAD-box RNA helicases have multiple functions and

4

have been involved in tumorigenesis.[18,19] However, as for DDX41, its function

5

has not been investigated until recently. A recent study reported that DDX41

6

cooperates with mRNA splicing factors and is involved in late spliceosomal

7

processing.[1] However, as we mentioned, we identified DDX41 p.R525H mutation

8

in AML exhibiting cytopenias in bone marrow and peripheral blood, whose

9

pathogenesis has not been known to be connected to mRNA splicing defect;

10

therefore, we further investigated the biological functions of DDX41 and its

11

contribution to the development of AML through an acquired mutation.

TE D

M AN U

SC

RI PT

1

The DDX41 localizes at the nuclear matrix and in the nucleolus, consistent

13

with the putative NLS at the N-terminus. However, opposite to our observation, a

14

recent study described DDX41 localization in the cytoplasm even when a full-length

15

cDNA was transfected.[6] The reason for the discrepancy between our observation

16

and others are not obvious, but it might be possible that a short form we identified

17

for the first time in this study was mainly expressed in previous studies describing

18

DDX41 in the cytoplasm. As it has also been reported that human DDX41 shuttles

AC C

EP

12

17

ACCEPTED MANUSCRIPT

1

between nucleus and cytoplasm,[20] it could be also possible that cells in which

2

DDX41 is localized mainly in the cytoplasm were used in previous studies. While the enforced expression of WT DDX41 in cord blood-derived cells had

4

minimal effects, substitution of p.R525 into histidine induced cell cycle arrest,

5

suggesting that this alteration interferes with cell growth in a dominant negative

6

manner. In addition, as expected, the inhibitory effect of p.R525H DDX41 on cell

7

cycle progression is due to the loss of the ATPase activity. Unexpectedly, the cell

8

cycle arrest was MDM2 and RB dependent, but p53 independent. However, this is

9

consistent

SC

previous

M AN U

with

works

arguing

that

MDM2-dependent

but

p53-independent pathways also cause ribosomopathies[21,22].

TE D

10

RI PT

3

Our results further indicate an interference of p.R525H DDX41 in

12

pre-rRNA processing, giving rise to a certain “ribosomal stress”. In humans, 18S,

13

5.8S, and 28S rRNA molecules are transcribed by RNA polymerase I as a single

14

precursor

15

post-transcriptionally into mature rRNA (Fig. S1B).[24] The processing of pre-rRNA

16

undergoes mainly in the nucleolus, where ~4,500 putative nucleolar proteins and

17

small nucleolar RNAs are supposed to participate in this process.[25] Although the

18

molecular function of almost all the nucleolar proteins has not been elucidated with

AC C

EP

11

called

47S

pre-rRNA[23];

18

it

is

then

modified

and

cleaved

ACCEPTED MANUSCRIPT

some exceptions, a recent study that performed siRNA-mediated depletion of 625

2

candidate nucleolar proteins helped to understand the pre-rRNA processing

3

machinery.[9] As it has been shown that some spliceosomal factors, such as

4

hPrp43/DHX15 RNA helicase, also participate in pre-rRNA processing,[26,27]

5

DDX41 could also play multiple roles. We think this pathway can at least partly

6

account for the development of AML with cytopenias harboring the DDX41 p.R525H

7

mutation. Hematopoietic stem cells (HSC) have a low level of protein synthesis

8

compared with differentiating or growing progenitor cells.[28] Assuming that AML

9

stem cells with the DDX41 p.R525H mutation are constitutively in a low protein

10

synthesis status due to a ribosomal stress, the cells could be able to be maintained

11

under this stress, but they cannot proliferate or differentiate, which would in part

12

explain the pathophysiology of a slowly-growing AML.

EP

TE D

M AN U

SC

RI PT

1

In summary, we propose a mechanism of growth defect in hematopoietic

14

cells triggered by p.R525H DDX41 occurring in the following order: (i) p.R525H

15

mutant inhibits pre-rRNA processing;, (ii) compromised ribosomal biogenesis as a

16

result of impaired rRNA synthesis causes a release of ribosomal proteins that bind

17

to MDM2; (iii) MDM2-mediated RB degradation is suppressed, thus eventually

18

activating the RB pathway and resulting in the inhibition of E2F activity. Although

AC C

13

19

ACCEPTED MANUSCRIPT

this study uncovered a pathogenic role of p.R525H DDX41 in the slow growth rate

2

of tumor cells, how the mutation induces AML development and inhibits cell

3

differentiation is still not understood. Lethally irradiated mice transplanted with

4

hematopoietic stem/progenitor cells over-expressing p.R525H DDX41 did not

5

develop myeloid malignancy, even in the p53-deficient background (Fig. S4B).

6

Considering late occurrence of AML in patients harboring the mutation, it might

7

require age-dependent epigenetic alterations or other somatic changes for this

8

mutation to fully transform hematopoietic cells.

9

M AN U

SC

RI PT

1

Conflict of interest disclosure

11

The authors declare no competing financial interests.

TE D

10

EP

12 Acknowledgements

14

We would like to thank Dr. H. Kawamoto for providing the cells and Mrs. R. Tai, E.

15

Kanai, and M. Nakamura for excellent technical assistance. This work was partly

16

supported by the Daiichi Sankyo Foundation of Life Science, The NOVARTIS

17

Foundation (Japan) for the Promotion of Science, the SGH Foundation, the Princess

18

Takamatsu Cancer Research Fund and Relay For Life program founded by the

AC C

13

20

ACCEPTED MANUSCRIPT

1

Japan Cancer Society.

3

References

4

1.

5

Polprasert C, Schulze I, Sekeres MA, et al. Inherited and Somatic Defects in DDX41 in Myeloid Neoplasms. Cancer Cell. 2015;27(5):658-670.

2.

SC

6

RI PT

2

Lewinsohn M, Brown AL, Weinel LM, et al. Novel germ line DDX41 mutations define families with a lower age of MDS/AML onset and lymphoid malignancies.

8

Blood. 2016;127(8):1017-1023.

9

3.

M AN U

7

Sekine R, Kitamura T, Tsuji T, Tojo A. Efficient retroviral transduction of human B-lymphoid and myeloid progenitors: marked inhibition of their growth

11

by the Pax5 transgene. Int J Hematol. 2008;87(4):351-362.

13

leukemia

14 15

Ding L, Ley TJ, Larson DE, et al. Clonal evolution in relapsed acute myeloid

EP

4.

revealed

by

whole-genome

sequencing.

Nature.

AC C

12

TE D

10

2012;481(7382):506-510.

5.

Parvatiyar K, Zhang Z, Teles RM, et al. The helicase DDX41 recognizes the

16

bacterial secondary messengers cyclic di-GMP and cyclic di-AMP to activate a

17

type I interferon immune response. Nat Immunol. 2012;13(12):1155-1161.

18

6.

Zhang Z, Yuan B, Bao M, Lu N, Kim T, Liu YJ. The helicase DDX41 senses

21

ACCEPTED MANUSCRIPT

1

intracellular DNA mediated by the adaptor STING in dendritic cells. Nat

2

Immunol. 2011;12(10):959-965. 7.

Subramanian A, Tamayo P, Mootha VK, et al. Gene set enrichment analysis: a

RI PT

3

knowledge-based approach for interpreting genome-wide expression profiles.

5

Proc Natl Acad Sci U S A. 2005;102(43):15545-15550.

6

8.

SC

4

Aspesi A, Pavesi E, Robotti E, et al. Dissecting the transcriptional phenotype of ribosomal protein deficiency: implications for Diamond-Blackfan Anemia. Gene.

8

2014;545(2):282-289.

9

9.

M AN U

7

Tafforeau L, Zorbas C, Langhendries JL, et al. The complexity of human ribosome biogenesis revealed by systematic nucleolar screening of Pre-rRNA

11

processing factors. Mol Cell. 2013;51(4):539-551.

EP

13

10. Linder P, Jankowsky E. From unwinding to clamping - the DEAD box RNA helicase family. Nat Rev Mol Cell Biol. 2011;12(8):505-516.

AC C

12

TE D

10

14

11. Dutt S, Narla A, Lin K, et al. Haploinsufficiency for ribosomal protein genes

15

causes selective activation of p53 in human erythroid progenitor cells. Blood.

16 17 18

2011;117(9):2567-2576. 12. Ebert BL, Pretz J, Bosco J, et al. Identification of RPS14 as a 5q- syndrome gene by RNA interference screen. Nature. 2008;451(7176):335-339.

22

ACCEPTED MANUSCRIPT

13. Fumagalli S, Ivanenkov VV, Teng T, Thomas G. Suprainduction of p53 by

2

disruption of 40S and 60S ribosome biogenesis leads to the activation of a novel

3

G2/M checkpoint. Genes Dev. 2012;26(10):1028-1040.

RI PT

1

14. Sun XX, Wang YG, Xirodimas DP, Dai MS. Perturbation of 60 S ribosomal

5

biogenesis results in ribosomal protein L5- and L11-dependent p53 activation. J

6

Biol Chem. 2010;285(33):25812-25821.

SC

4

15. Durland-Busbice S, Reisman D. Lack of p53 expression in human myeloid

8

leukemias is not due to mutations in transcriptional regulatory regions of the

9

gene. Leukemia. 2002;16(10):2165-2167.

M AN U

7

16. Sato S, Li K, Kameyama T, et al. The RNA sensor RIG-I dually functions as an

11

innate sensor and direct antiviral factor for hepatitis B virus. Immunity.

12

2015;42(1):123-132.

14 15 16 17 18

EP

17. Jiang F, Ramanathan A, Miller MT, et al. Structural basis of RNA recognition

AC C

13

TE D

10

and

activation

by

innate

immune

receptor

RIG-I.

Nature.

2011;479(7373):423-427.

18. Fuller-Pace FV. DEAD box RNA helicase functions in cancer. RNA Biol. 2013;10(1):121-132. 19. Dardenne E, Pierredon S, Driouch K, et al. Splicing switch of an epigenetic

23

ACCEPTED MANUSCRIPT

1

regulator by RNA helicases promotes tumor-cell invasiveness. Nat Struct Mol

2

Biol. 2012;19(11):1139-1146.

4

20. Abdul-Ghani M, Hartman KL, Ngsee JK. Abstrakt interacts with and regulates

RI PT

3

the expression of sorting nexin-2. J Cell Physiol. 2005;204(1):210-218.

21. Donati G, Brighenti E, Vici M, et al. Selective inhibition of rRNA transcription

6

downregulates E2F-1: a new p53-independent mechanism linking cell growth

7

to cell proliferation. J Cell Sci. 2011;124(Pt 17):3017-3028.

M AN U

SC

5

8

22. Uchida C, Miwa S, Kitagawa K, et al. Enhanced Mdm2 activity inhibits pRB

9

function via ubiquitin-dependent degradation. EMBO J. 2005;24(1):160-169.

13

TE D

12

Blood Rev. 1996;10(3):185-188.

24. Mullineux ST, Lafontaine DL. Mapping the cleavage sites on mammalian

EP

11

23. Baatout S. Staining of the nucleolar organizer regions: relevance in hematology.

pre-rRNAs: where do we stand? Biochimie. 2012;94(7):1521-1532.

AC C

10

14

25. Ahmad Y, Boisvert FM, Gregor P, Cobley A, Lamond AI. NOPdb: Nucleolar

15

Proteome Database--2008 update. Nucleic Acids Res. 2009;37(Database

16

issue):D181-184.

17

26. Bohnsack MT, Martin R, Granneman S, Ruprecht M, Schleiff E, Tollervey D.

18

Prp43 bound at different sites on the pre-rRNA performs distinct functions in

24

ACCEPTED MANUSCRIPT

1

ribosome synthesis. Mol Cell. 2009;36(4):583-592. 27. Yoshimoto R, Okawa K, Yoshida M, Ohno M, Kataoka N. Identification of a

3

novel component C2ORF3 in the lariat-intron complex: lack of C2ORF3

4

interferes with pre-mRNA splicing via intron turnover pathway. Genes Cells.

5

2014;19(1):78-87.

SC

7

28. Buszczak M, Signer RA, Morrison SJ. Cellular differences in protein synthesis regulate tissue homeostasis. Cell. 2014;159(2):242-251.

M AN U

6

RI PT

2

8 9

Figure legends

Figure 1. Identification of the somatic DDX41 p.R525H mutation in AML patients

11

exhibiting cytopenias.

12

(A) Bone marrow blasts of three AML patients harboring the p.R525H DDX41

14

EP

mutation.

The

images

were

taken

at

400x

magnification.

Clinical

AC C

13

TE D

10

manifestations of the patients are shown in the table at the bottom.

15

(B) Confirmation of somatic DDX41 p.R525H mutation by Sanger re-sequencing.

16

Representative data of CD34-positive, CD34-negative/CD3-negative, and

17

CD3-positive cells from the patient UPN17 are shown. An arrow indicates the

18

p.R525 position.

25

ACCEPTED MANUSCRIPT

1 Figure 2. Nuclear localization of DDX41.

3

(A) Localization of Myc-tagged human DDX41 protein in murine fibroblasts by

4

immunofluorescent analysis. DDX41 protein with or without p.R525H

5

substitution was expressed together with a FLAG-tagged STING. An anti-Myc

6

rabbit antibody and an anti-FLAG mouse antibody were used as primary

7

antibodies; a Cy3-labelled anti-rabbit antibody and an Alexa488-labelled

8

anti-mouse antibodies were used as secondary antibodies. Nuclei were stained

9

with Hoechst 33342. Scale bars: 20µM.

M AN U

SC

RI PT

2

(B) Endogenous DDX41 expression in THP-1 cells. The cells attached onto a

11

slide-glass using a cytospin apparatus were stained with an anti-DDX41 and an

12

anti-Nucleolin antibody, followed by the staining with a Cy3-labelled anti-rabbit

13

antibody and an Alexa488-labelled anti-mouse antibody, respectively. The

15

EP

AC C

14

TE D

10

images were taken using a LSM-5 PASCAL confocal microscopy. Scale bar: 20µM.

16

(C) A schematic illustration of the DDX41 protein structure. The helicase core

17

domain is shown in light blue. A predicted NLS, the second methionine, and the

18

motif VI in the helicase domain are shown in red characters.

26

ACCEPTED MANUSCRIPT

1 Figure 3. Cell growth inhibition by DDX41 p.R525H mutation that causes impaired

3

ribosomal biogenesis.

4

(A) Proliferation

of

CD34-positive

cord

blood

cells

RI PT

2

transfected

with

a

DDX41-expressing vector or an empty vector in the presence of SCF, TPO, and

6

FLT3L. The left graph indicates the ratio between the number of cells at 30

7

days and at the beginning of the assay. The right graph shows DDX41 mRNA

8

expression levels as indicated by rpkm values.

M AN U

SC

5

(B) ATPase assay using a helicase domain of DDX41 (WT and p.R525H mutant)

10

protein as catalyst. In vitro translated helicase domains were mixed with

11

[α-32P]ATP for the indicated minutes, and the mixtures were subjected to

12

thin-layer chromatography. Arrows indicate ATP and ADP.

14 15

EP

(C) Detection of pre-rRNA intermediates expressed in THP-1 cells by Northern

AC C

13

TE D

9

blotting analysis. DIG-labelled RNA probes for ITS1 and ITS2 regions were used for the detection (Fig. S1A). The right panels indicate lane profile plots

16

generated from the Northern blots (left panels) using Image J software and

17

quantitative values of each pre-rRNA band relative to the empty vector control.

18

(D) Relative amounts of pre-rRNA spacers as assessed by qPCR analysis. Empty

27

ACCEPTED MANUSCRIPT

1

vector or DDX41 (p.R525H and WT)-transfected cord blood cells were subjected

2

to qPCR experiments. The regions amplified by PCR are shown in Fig. S1A.

RI PT

3

Figure 4. Ribosomal stress that results in the activation of RB pathway occurring in

5

p.R525H expressing cells.

6

(A) Negative enrichment of cell cycle-promoting genes regulated by the RB-E2F axis

7

in DDX41 p.R525H cord blood cells. GSEA analyses were performed using

8

mRNA sequencing data, in which the plots of “Chang Cycling Genes”, “Hallmark

9

E2F Targets”, and “Eguchi Cell Cycle RB1 Targets” are shown as

10

representatives. Normalized Enrichment Score, nominal p-value and False

11

Discovery Rate (FDR) q-value are indicated at the bottom.

TE D

M AN U

SC

4

(B) Negative enrichment of cell cycle-promoting genes regulated by the RB-E2F axis

13

in CD34-positive tumor cells obtained from patients (3 patients with the

15

AC C

14

EP

12

p.R525H mutation were compared with 20 patients without the mutation). GSEA analyses were performed as in (A).

16

(C) Increased binding of RPL5 and RPL11 to MDM2 in p.R525H DDX41-expressing

17

cells. Total protein extracts from K562 cells expressing DDX41 (WT or p.R525H)

18

or

those

transfected

with

an

28

empty

vector

were

subjected

to

ACCEPTED MANUSCRIPT

1

immunoprecipitation analysis using an anti-MDM2 antibody or a control IgG,

2

followed by the detection of RPL5, RPL11 and MDM2 proteins. (D) The expression of RB and its phosphorylated and poly-ubiquitinated forms, of

4

MDM2, Myc-tagged DDX41, actin and poly-ubiquitinated proteins in K562 cells

5

transfected with an empty vector or with a DDX41 (WT or p.R525H) vector. In

6

the immunoprecipitation analysis (right panels indicated as IP), proteins

7

precipitated

8

anti-Multi-Ubiquitin antibody and an anti-RB antibody.

anti-RB

antibody

were

M AN U

an

AC C

EP

TE D

with

SC

RI PT

3

29

blotted

using

an

ACCEPTED MANUSCRIPT Figure 1 A.

UPN10

UPN6

UPN17

Age/Sex

Karyotype

WBC (103/µ µL)

Blast(%) (PB)

Blast(%) (BM)

Hb (g/dL)

Plt (104/µ µL)

Other gene mutations*

6

72/M

45,X,-Y (8/20cells), 46,XY (12/20cells)

1.6

1

25

11.3

6.7

SF1

10

81/M

47,XY,+der(1;19)(q10;p10) (3/20cells), 46,XY (17/20cells)

0.6

1

32

6.4

1.6

ー

17

64/M

46,XY

3.1

3.3

28

4.8

16.6

ー

2995 (TCGA)

67/M

46,XY

0.6

0

35

N.I.

N.I.

MYLK2

SC

p.R525 (c.G1574)

Patient: UPN17 CD34+ cells

Amino acid V

H

R

I

G

R

T

G R/H S

G N T

G

I

A

R

I

G

R

T

G R/H S

G

G

I

A

CD3+ cells

H

N T

TE D

CD34-CD3- cells

Amino acid V

M AN U

*mutations previously reported as recurrent. N.I.: not indicated.

B.

EP

nucleotide GTACAC CGG ATT GGC CGC ACC GGG CGC TCG GGA AACACAGGC ATC GCC amino acid V H R I G R T G R S G N T G I A

AC C

RI PT

UPN

ACCEPTED MANUSCRIPT Figure 2 A.

Anti-FLAG (Alexa488)

merged

Anti-Myc (Cy3)

B. Merged

Hoechst33342

Anti-DDX41 (Cy3)

DDX41 (WT)-Myc FLAG-STING

RI PT

DDX41 (p.R525H)-Myc FLAG-STING

AntiNucleolin (Alexa488)

DDX41 N

6 ↓ PERKRARTD Predicted NLS

127 ↓ M

M AN U

C.

SC

FLAG-STING

Hoechst33342

200 ↓

525 ↓ YVHRIGRTGRSGN Motif VI Helicase core domain

2nd methionine

AC C

EP

TE D

52-kDa 70-kDa

622 (Amino acid) ↓ C

ACCEPTED MANUSCRIPT Figure 3 Cell proliferation (Ratio between 30 days and 0 day) 60 40 20 0 Empty vector

WT

B. DDX41 mRNA

(rpkm) 300 200 100 0

ATP Empty vector

p.R525H

ADP

WT

p.R525H

BSA

p.R525H WT p.R525H

DDX41 (helicase domain)

DDX41

DDX41

WT

RI PT

A.

30min.

15min.

C.

DDX41 Empty vector WT p.R525H

DDX41 Empty vector WT p.R525H

47S

47S 41S

47/45S

41S

41S

1.00

1.00

WT 0.84

M AN U 21S

26S

0.94

1.00

30S

32S

30S

DDX41

SC

ITS1 probe

Empty vector

1.00

0.92 0.85

p.R525H 1.38 1.36 1.21 0.27

ITS2 probe

47S/45S 1.00 41S 1.00 1.00 32S

12S

ITS2 probe

EP

ITS1 probe

TE D

21S

D.

Relative expression of pre-rRNA spacers 4 3 2 1

AC C

5

0 5'-01

01-A0

A0-1

3-E

5'ETS Empty vector

E-C

2-5.8S

ITS1 DDX41 WT

DDX41 p.R525H

4-4a ITS2

12S

1.00

0.95 1.81

0.95

0.94 0.95 1.55

1.80

2.26

ACCEPTED MANUSCRIPT Figure 4

-0.2

-0.2

-0.4

-0.4

0.0 -0.2 -0.4 -0.6

-0.6

-0.6

Cord blood DDX41 p.R525H

DDX41 WT

-0.8 DDX41 p.R525H

-0.2

-0.2

-0.4

-0.4

-0.6

-0.6

DDX41 WT

Eguchi Cell Cycle RB1 Targets -2.04 0.0 0.0

Eguchi Cell Cycle RB1 Targets 0.0

-0.2 -0.4 -0.6

M AN U

Enrichment Score

0.0

0.0

Without With DDX41 mutation DDX41 mutation

Normalized Enrichment Score Nominal p-value FDR q-value

C.

Hallmark E2F Targets -2.46 0.0 0.0

Hallmark E2F Targets

Chang Cycling Genes

Patients With DDX41 mutation

DDX41 p.R525H

DDX41 WT

Chang Cycling Genes -2.14 0.0 0.0

Normalized Enrichment Score Nominal p-value FDR q-value

B.

Eguchi Cell Cycle RB1 Targets

Hallmark E2F Targets 0.0

RI PT

Enrichment Score

Chang Cycling Genes 0.0

SC

A.

Without With DDX41 mutation DDX41 mutation

Chang Cycling Genes -2.30 0.0 0.0

Hallmark E2F Targets -2.19 0.0 0.0

Without DDX41 mutation

Eguchi Cell Cycle RB1 Targets -1.80 0.0 0.02

TE D

IP: Control IgG Input IP: Anti-MDM2 DDX41 DDX41 DDX41 Empty Empty Empty vector WT p.R525H vector vector WT p.R525H WT p.R525H

D.

AC C

AntiRPL11 AntiMDM2

Short-time exposure Long-time exposure

EP

AntiRPL5

Empty vector

Anti-RB Antiphosho-RB Anti-MDM2

Short-time exposure Long-time exposure

DDX41

WT

Empty vector

p.R525H AntiMulti-Ub

DDX41 WT

IP: Anti-RB

p.R525H (KDa) 150 102

Empty vector AntiMulti-Ub

76

Anti-Myc (70-kDa DDX41)

52

Anti-Actin

38

Anti-RB

DDX41 WT

p.R525H

ACCEPTED MANUSCRIPT

Supplemental Figure legends

2

Figure S1. Structure and processing pathways of pre-rRNA.

3

(A) The structure of 47S pre-rRNA. The 5’ and 3’ ETS and ITS 1 and 2 are spacer

4

regions that will be removed during pre-rRNA processing, and there are many

5

pre-rRNA intermediates partially retaining these spacers as shown in Fig. S1B.

6

Detailed pre-rRNA cleavage sites and their names are described in a recent

7

review (Mullineux ST and Lafontaine DL, Biochimie. 2012; 94, 1521-1532). The

8

regions used for Northern probes (ITS1 and ITS2) and amplified by qPCR

9

analysis (5’-01, 01-A0, A0-1, 3-E, E-C, 2-5.8S and 4’-4a) are shown in orange and

SC

M AN U

gray, respectively.

TE D

10

RI PT

1

(B) A schematic processing pathways of pre-rRNA. Three rRNAs (18S, 5.8S and

12

28S) are first transcribed as a long polycistronic precursor designated as

13

47S pre-rRNA, which will be subsequently processed to produce mature

15

AC C

14

EP

11

rRNAs through pathways as shown in this schema.

16

Figure S2. Expression and localization of DDX41 and of its short isoform.

17

(A) Endogenous DDX41 expression in four myeloid leukemia cell lines (upper

18

panels). DDX41 was detected by an antibody against the C-terminus of DDX41.

1

ACCEPTED MANUSCRIPT

1

A full-length image of DDX41 immunoblot is shown at right. (B) Exogenous DDX41 proteins expressed in HEK293 cells. The cells were

3

transfected with a Myc-tagged DDX41 cDNA (cDNA starting from the 1st

4

methionine and from the 2nd methionine, designated as p70 and p52,

5

respectively), and the extracts were subjected to immunoblot analysis. Tagging

6

positions are indicated as C (C-terminus) or N (N-terminus). The C-terminally

7

tagged p70 cDNA also showed the 52-kDa band weakly (lanes 1 and 2), whereas

8

the p70 cDNA tagged at the N-terminus did not show the 52-kDa band (lane 5

9

and 6), which suggests that the 52-kDa band is a truncated protein originated

10

from the c-terminal part of DDX41. Indeed, the p70 cDNA where the p.M132

11

(the 3rd methionine) was substituted with an alanine expressed a very weak

12

52-kDa band (lane 11), which disappeared when the p.M127 (the 2nd methionine)

13

was substituted to alanine (lane 13 and 14). Actin was used as a loading control.

15 16 17

SC

M AN U

TE D

EP

AC C

14

RI PT

2

A schematic diagram was shown at the middle, and full-length images of DDX41 immunoblot are shown at bottom right.

(C) The localization of 52-kDa DDX41 tagged with Myc at its c-terminus. The experiment was performed as in Fig. 2A. Scale bars: 20µM.

18 2

ACCEPTED MANUSCRIPT

Figure S3. Suppression of ribosomal genes in p.R525H DDX41 expressing cord

2

blood cells.

3

(A) Decrease in ribosomal genes (RPLs and RPSs) in p.R525H DDX41 cord blood

4

cells compared with the cells transfected with wildtype DDX41-expressing

5

vector (WT cells). Each bar indicates the ratio of p.R525H cells to WT cells.

SC

RI PT

1

(B) GSEA analysis indicating negative enrichment of gene sets related to ribosome

7

biogenesis and positive enrichment of genes upregulated in RPS14-deficient in

8

p.R525H DDX41 cord blood cells. Normalized Enrichment Score, nominal

9

p-value and FDR q-value are shown at the bottom.

10

TE D

M AN U

6

Figure S4. No growth advantage of p.R525H DDX41 expressing cells in vitro and in

12

vivo.

13

(A) DDX41 (WT or p.R525H)-expressing THP-1 cell number. THP-cells lacking

15

AC C

14

EP

11

functional p53 were transduced with a DDX41 expressing vector and viable cell numbers were counted.

16

(B) No growth advantage of p.R525H DDX41-expressing hematopoietic cells in the

17

murine BMT experiments. Donor cells (with or without p53+/- background) were

18

transduced with an empty/GFP-expressing vector or a DDX41/GFP-expressing

3

ACCEPTED MANUSCRIPT

vector, followed by the transplantation into lethally irradiated recipients. GFP

2

positivity (%) and peripheral WBC numbers (x103/µL) were measured two to

3

three months after the BMT. No significant difference in GFP signals was

4

observed between the groups, suggesting that enforced expression of DDX41

5

pR525H alone does not account for the acquisition of clonogenicity.

AC C

EP

TE D

M AN U

SC

RI PT

1

4

ACCEPTED MANUSCRIPT Figure S1 A.

01 A0

ITS2 region

ITS1 region

5’ETS region

1

3EC 2

4’

4a 3’

3’

5’ qPCR amplicons

3

01 A0

01

5.8S

A0

ITS2

5.8S 4’

4a

SC

1

ITS1 E E C 2

28S

B. 47S 5’ end

02

Pathway 1 02

7S

02 32S

5.8S

AC C

3

12S

30S 01

TE D

E

18S-E 1

2

3’

4a

EP

C

21S-C 1

18S 1

Pathway 2

1

2

M AN U

3’ end

45S 01

1

RI PT

18S Northern probes

21S

02

5’

47S pre-rRNA

41S

3’ETS region

4’

28S

21S 1

18S-E 1

18S 1

2

2

12S

3’

7S

4a

C

21S-C 1

E

3

02 32S

2

5.8S

4’

28S

ACCEPTED MANUSCRIPT Figure S2 A. F-36P

Cell line

FKH-1

HEL

Full-length image of the upper left panel

K562

70-kDa AntiDDX41 52-kDa

p70

B.

p52

p70

p70

Empty WT p.R525H WT p.R525H WT p.R525H vector Myc-tag (C- or N-terminus) C C C C N N

52-kDa

1

2

3

4

5

DDX41 Myc-tag

N 52-kDa

C

2nd methionine (M127) st 1 methionine (M1)

70-kDa

N

EP merged

52-kDa DDX41 (p.R525H)-Myc FLAG-STING

10

11

12

FLAG (Alexa488) Myc (Cy3)

13

DDX41 Myc-tag N

70-kDa

C Anti-Myc Ab

52-kDa C

Detection of only 70-kDa DDX41 (Lanes 5 and 6)

52-kDa DDX41 (WT)-Myc FLAG-STING

C

2nd methionine (M127) st 1 methionine (M1)

AC C

C.

Anti-Myc Ab

9

DDX41

N

Anti-Myc Ab

8

7

Myc-tag

C

Detection of 70-kDa and 52-kDa DDX41 by anti-Myc Ab (Lanes 1, 2 and 8)

6

TE D

70-kDa

p70 P70 (p.M132A) (p.M127A/M132A)

SC

70-kDa

Anti-Actin

N

p52

WT WT p.R525H WT p.R525H C C C C C

M AN U

Anti-Myc (DDX41)

WT C

RI PT

Anti-Actin

M127A/M132A 1st methionine (M1)

Expression of only 70-kDa DDX41 (Lanes 12 and 13)

Full-length images of the upper panels

Hoechst33342

ACCEPTED MANUSCRIPT Figure S3 A. 2

Relative RPL gene expression (Ratio: DDX41 p.R525H to WT)

1.5 1

0 RPL 1

2

3

4

5

6

7

8

RI PT

0.5

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

2

Relative RPS gene expression (Ratio: DDX41 p.R525H to WT)

SC

1.5 1

2

4

5

6

7

8

9

10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Ribonucleoprotein_Complex Biogenesis_and_Assembly 0.0

KEGG_Ribosome 0.0 -0.2

-0.2

-0.4 -0.6

TE D

-0.4 -0.6

DDX41 DDX41 DDX41 DDX41 WT p.R525H WT p.R525H Hallmark Myc Targets V1 RPS14_DN.V1_UP 0.5 0.0 0.4 -0.2 0.3 0.2 -0.4 0.1 -0.6 0.0

EP

Enrichment Score

Enrichment Score

B.

3

AC C

0 RPS 1

M AN U

0.5

DDX41 p.R525H

Normalized Enrichment Score Nominal p-value FDR q-value

DDX41 DDX41 WT p.R525H

DDX41 WT

Ribonucleoptotein Complex Biogenesis and Assembly

KEGG Ribosome

Hallmark Myc Targets V1

RPS14 DN.V1_UP

-1.94

-1.67

-2.28

1.73

0.0 0.006

0.003 0.008

0.0 0.0

0.0 0.001

ACCEPTED MANUSCRIPT Figure S4 A. (x105/mL)

Cell count

6 5 DDX41 WT 4

DDX41p.R525H

RI PT

3 2 1 0 4

5

6

7

9 10 11 12 13 (Day)

8

GFP positivity

(x103/µL) 20

15

15

10

10

5

5

0

0 Empty vector

WT

p.R525H p.R525H (p53+/-) DDX41

WBC count

Empty vector

WT

p.R525H p.R525H (p53+/-)

TE D

DDX41

EP

SC

(%) 20

3

AC C

B.

2

M AN U

0