blood-2011-charfi-1899-910 ppo

8/21/2019 Blood-2011-Charfi-1899-910 ppo

1/13

doi:10.1182/blood-2010-10-311001Prepublished online December 6, 2010;2011 117: 1899-1910

Cyndia Charfi, Véronique Voisin, Louis-Charles Levros, Jr, Elsy Edouard and Eric Rassart oncogeneleukemias: identification of leukemia markers and Fmn2 as a potentialGene profiling of Graffi murine leukemia virus-induced lymphoid

http://bloodjournal.hematologylibrary.org/content/117/6/1899.full.htmlUpdated information and services can be found at:

(1217 articles)Lymphoid Neoplasia Articles on similar topics can be found in the following Blood collections

http://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#repub_requestsInformation about reproducing this article in parts or in its entirety may be found online at:

http://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#reprintsInformation about ordering reprints may be found online at:

http://bloodjournal.hematologylibrary.org/site/subscriptions/index.xhtmlInformation about subscriptions and ASH membership may be found online at:

Copyright 2011 by The American Society of Hematology; all rights reserved.Washington DC 20036.by the American Society of Hematology, 2021 L St, NW, Suite 900,Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly

For personal use only.by guest on October 5, 2012. bloodjournal.hematologylibrary.orgFrom

http://bloodjournal.hematologylibrary.org/content/117/6/1899.full.htmlhttp://bloodjournal.hematologylibrary.org/content/117/6/1899.full.htmlhttp://bloodjournal.hematologylibrary.org/content/117/6/1899.full.htmlhttp://bloodjournal.hematologylibrary.org/cgi/collection/lymphoid_neoplasiahttp://bloodjournal.hematologylibrary.org/cgi/collection/lymphoid_neoplasiahttp://bloodjournal.hematologylibrary.org/cgi/collection/lymphoid_neoplasiahttp://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#repub_requestshttp://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#repub_requestshttp://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#reprintshttp://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#reprintshttp://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#reprintshttp://bloodjournal.hematologylibrary.org/site/subscriptions/index.xhtmlhttp://bloodjournal.hematologylibrary.org/site/subscriptions/index.xhtmlhttp://bloodjournal.hematologylibrary.org/site/subscriptions/index.xhtmlhttp://bloodjournal.hematologylibrary.org/subscriptions/ToS.dtlhttp://bloodjournal.hematologylibrary.org/subscriptions/ToS.dtlhttp://bloodjournal.hematologylibrary.org/subscriptions/ToS.dtlhttp://bloodjournal.hematologylibrary.org/subscriptions/ToS.dtlhttp://bloodjournal.hematologylibrary.org/subscriptions/ToS.dtlhttp://bloodjournal.hematologylibrary.org/http://bloodjournal.hematologylibrary.org/http://bloodjournal.hematologylibrary.org/subscriptions/ToS.dtlhttp://bloodjournal.hematologylibrary.org/http://bloodjournal.hematologylibrary.org/subscriptions/ToS.dtlhttp://bloodjournal.hematologylibrary.org/site/subscriptions/index.xhtmlhttp://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#reprintshttp://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#repub_requestshttp://bloodjournal.hematologylibrary.org/cgi/collection/lymphoid_neoplasiahttp://bloodjournal.hematologylibrary.org/content/117/6/1899.full.html

8/21/2019 Blood-2011-Charfi-1899-910 ppo

2/13

LYMPHOID NEOPLASIA

Gene profiling of Graffi murine leukemia virus-induced lymphoid leukemias:identification of leukemia markers and Fmn2 as a potential oncogene

Cyndia Charfi,1 Véronique Voisin,1 Louis-Charles Levros, Jr,1 Elsy Edouard,1 and Eric Rassart1

1Laboratoir e de Biologie Moléculaire, Départemen t des Sciences Biologique s, Centre BioMed, Univers ité du Québec à Montréal, M ontréal, Q C

The Graffi murine leukemia virus induces

a large spectrum of leukemias in mice

and thus provides a good model to com-

pare the transcriptome of all types of

leukemias. We analyzed the gene expres-

sion profiles of both T and B leukemias

induced by the virus with DNA microar-

rays. Given that we considered that a

4-fold change in expression level was

significant, 388 probe sets were associ-

ated to B, to T, or common to both leuke-

mias. Several of them were not yet associ-

ated with lymphoid leukemia. We confirmed

specific deregulation of Fmn2 , Arntl2 ,

Bfsp2 , Gfra2 , Gpm6a , and Gpm6b in B

leukemia, of Nln , Fbln1, and Bmp7 in T

leukemias, and of Etv5 in both leukemias.

More importantly, we show that themouse

Fmn2 induced an anchorage-independent

growth, a drastic modification in cell

shape with a concomitant disruption of

the actin cytoskeleton. Interestingly, we

found that human FMN2 is overexpressed

in approximately 95% of pre-B acute lym-

phoblastic leukemia with the highest

expression levels in patients with a TEL /

AML1 rearrangement. These results,

surely related to the role of FMN2 in

meiotic spindle maintenance, suggest its

important role in leukemogenesis. Fi-

nally, we propose a new panel of genes

potentially involved in T and/or B leuke-

mias. (Blood . 2011;117(6):1899-1910)

Introduction

To understand the mechanism of induction of leukemia in humans,

the mouse has been considered to be an ideal model given the

extent of genetic similarity between these 2 species. We have

already shown that the Graffi murine leukemia virus (MuLV), when

inoculated into newborn mice, induces a broad spectrum of

leukemias of lymphoid (T or B) and nonlymphoid (myeloid,

erythroid, or megakaryoblastic) origins.1 We also demonstrated

that the Graffi MuLV directly targets the c-Myc, Fli1, Pim1, and

Spi1 / P1 oncogenes2 but also the Gris1 locus that encodes an

oncogenic truncated form of cyclin D2.3,4 We took advantage of the

large spectrum of leukemias induced by the Graffi murine leukemia

virus to analyze and compare the transcriptome of these leukemiaswith a DNA microarray approach. One clear advantage of microar-

ray analysis is that genes that are not targeted through retroviral

integration but act as oncogene on deregulation can be equally

identified. Using this approach, we recently identified several genes

that are directly involved in erythroid and megakaryoblastic

leukemias in both mouse and human.5 In this report, we have

specifically compared the transcriptomes of T and B lymphoid

leukemias induced by Graffi MuLV with their corresponding

controls. We identified new relevant signatures that highlight many

potential markers or oncogenes for T and B leukemias (especially

for pre-B-leukemia), some of which were common to both types of

lymphoid leukemia. Among the selected genes, we validated the

modulation of the expression of 10 genes of 12, the functions of which remained poorly elucidated in lymphoid leukemias. Further-

more, for the first time, we provide data supporting a role for Fmn2,

a member of the formin family in leukemogenesis in pre-B-lineage

acute lymphoblastic leukemia (B-ALL) and particularly, in pediat-

ric pre-B-ALL harboring the t(12;21) TEL/AML1 translocation.

Methods

Human sample collection

For this study, samples from 12 pediatric pre-B-ALL were obtained from Dr

Daniel Sinnet (Sainte-Justine Hospital, Montreal, QC), whereas samples

from 13 adult patients with different types of B leukemia were obtained

from the Quebec Leukemia Cell Bank. As a control (CH), we pooled

peripheral blood mononuclear cells isolated from 10 healthy adults.

Information relative to each patient is provided in supplemental Table 1

(available on the Blood Web site; see the Supplemental Materials link at the

top of the online article). The research protocol was approved by Ethic

Committees of all concerned institutions.

Mice sample collection

Newborn ( 24 hours) NFS mice were injected intraperitoneally with viral

particles of Graffi MuLV variant GV-1.4 (1 106 PFU) or GV-1.2

(3 106 PFU).1 Lymph nodes, thymus, bone marrows, and spleens were

harvested from moribund mice for flow cytometry analysis and RNA

extraction. All the experimental procedures were approved by the Animal

Care Committee of the University of Quebec (Montreal, QC).

Flow cytometry and cell isolation

Flow cytometry was performed as previously described.1 Cell populations

(tumor or control) were purified from the hematopoietic organs by positive

selection using magnetic beads coated with the chosen antibody. Leukemic

cells were sorted as follow: T cells from the thymus, B cells from the

enlarged lymph nodes, and erythroid and megakaryoblastic cells from the

infiltrated spleen. Nonleukemic control cells were sorted from a pool of

12 noninfected NFS mice: T cells from the thymus, B cells from the spleen,

and erythroblasts and megakaryoblasts from the bone marrow.

Submitted September 30, 2010; accepted November 16, 2010. Prepublished

onlineas Blood First Editionpaper, December 6, 2010; DOI 10.1182/blood-2010-10-

311001.

The online version of this article contains a data supplement.

The publication costs of this article were defrayed in part by page charge

payment. Therefore, and solely to indicate this fact, this article is hereby

marked ‘‘advertisement’’ in accordance with 18 USC section 1734.

© 2011 by The American Society of Hematology

1899BLOOD, 10 FEBRUARY 2011 VOLUME 117, NUMBER 6


http://bloodjournal.hematologylibrary.org/subscriptions/ToS.dtlhttp://bloodjournal.hematologylibrary.org/subscriptions/ToS.dtlhttp://bloodjournal.hematologylibrary.org/http://bloodjournal.hematologylibrary.org/http://bloodjournal.hematologylibrary.org/subscriptions/ToS.dtlhttp://bloodjournal.hematologylibrary.org/

8/21/2019 Blood-2011-Charfi-1899-910 ppo

3/13

RNA isolation, microarray hybridization, and

dataset normalization

Total RNA was directly isolated from spleen, thymus, lymph nodes, and

bone marrow samples or after cell sorting with the Trizol reagent

(Invitrogen) and purified using RNeasy Mini Kit (QIAGEN). Total RNA

(2 g) from each sample was prepared for hybridization to Affymetrix

Gene Chip Mouse Genome 430, Version 2.0 arrays (Genome Quebec

Innovation Center). The Affymetrix MicroArray Suite, Version 5.0 wasused to scan and quantify the arrays. Normalization of gene expression data

was performed using the Bioconductor implementation of Robust Multi

Array (B. Bolstad, University of California, Berkeley, CA) available from

the Flexarray software Version 1.2, R 2.7.2.6

Microarray data analyses

Using the gene expression profiles obtained from selected leukemias, the

Robust Multi Array values of the 45 000 probe sets were used to identify

differentially expressed genes. Genes with signal intensity significantly

higher (up-regulated) or lower (down-regulated) in leukemias versus

control cells (ie, with a 4-fold change in expression levels) were selected.

To group microarrays and/or genes based on the high degree of their

expression patterns, hierarchical clustering (complete linkage clustering,

correlation uncentered) was constructed using GeneCluster software Ver-sion 1.6.7 Probe sets selected were further analyzed and the results were

treated with the Tree View program. The NetAffx website (Affymetrix) was

also used to retrieve gene ontology annotations, probe sequences, and used

as a link to Unigene (National Center for Biotechnology Information) for

further functional studies. The microarray dataset was deposited at Gene

Expression Omnibus under the accession number GSE12581.

Semiquantitative RT-PCR

Reverse transcription (RT) reactions were performed with oligo(dT) as

primer using the Omniscript enzyme (QIAGEN) and 100 ng of total RNA.

Using an RT reaction corresponding to 10 ng of tumor RNA samples for

each selected gene and to 2 ng for actin, the polymerase chain reactions

(PCRs) were performed with the Taq polymerase kit (QIAGEN) and thefollowing conditions: 94°C for 3 minutes, 94°C for 45 seconds, 56°C for

45 seconds, 72°C for 30 seconds, with a final extension at 72°C for

10 minutes. Annealing temperature and number of cycles were optimized

for each of the selected 12 genes with specific primers (supplemental Table

2) for semiquantitative analysis. PCR products were analyzed on agarose

gel, and band density was quantified with Quantity One Image Software

Version 4.4, using the actin gene as an internal control.

Cellular localization of Fmn2

NIH/3T3 fibroblasts, obtained from ATCC, were grown in Dulbecco

modified Eagle medium supplemented with 10% calf serum, 50 U penicil-

lin, and 50 g of streptomycin (Invitrogen), 16 hours before transfection.

Cells were, respectively, transfected with p-EGFP-N1 (control vector) and

GFP-tagged Fmn2 using the polyfect reagent (QIAGEN). Fmn2 localiza-tion was analyzed by confocal microscopy 48 hours after transfection. For

colocalization, transfected cells were plated on glass coverslips and grown

at 50% confluence. Colocalization with the cell surface membrane was

determined after cell staining for 5 minutes with 2.5 g/mL of CellMask

Plasma Membrane Stains (Invitrogen) and washing with phosphate-

buffered saline (PBS). Actin filament staining was performed after cell

fixation for 20 minutes with 4% paraformaldehyde, followed by PBS

washes and permeabilization for 5 minutes with 0.1% Triton X-100 in PBS.

Cells were incubated 1 hour in PBS with 1% bovine serum albumin, washed

twice with PBS and the coverslips, and then incubated with 0.3M of

phalloidin coupled to the AlexaFluor-555 (Invitrogen) for 20 minutes. After

2 washes with PBS, coverslips were mounted onto slides using Prolong

Gold antifade reagent (Invitrogen) and observed within 24 hours by

confocal microscopy. For -tubulin staining, the primary antibody used was

a mouse monoclonal anti--tubulin (1:2000; Sigma-Aldrich).

Colony formation in soft agar

To determine the anchorage-independent growth, colony formation was

tested in soft agar as previously described.8 Briefly, NIH/3T3 cells were

transiently transfected with 2.5 g of empty vector (pCMV), a Ras EJ 6.6,

or pCMV-Fmn2 expression vector. After 48 hours, 1 104 cells were

mixed with melted 0.3% agarose in Dulbecco modified Eagle medium and

seeded in 6-well plates on top of a 0.6% agarose base layer containing the

same medium. The top layer was covered with 1.5 mL of Dulbeccomodified Eagle medium. Cells were fed twice a week for 4 weeks and

observed with an optical microscope (40, Ernst Leitz, 6MBH Wetzlar).

Colonies whose size was at least twice larger than that of control colonies

were counted.

Results

To better elucidate the cancer signatures of B and T leukemias

induced by the murine Graffi virus and to identify new oncogene

candidates, a microarray analysis was performed on different types

of B and T leukemias induced by this virus compared with

nonleukemic B-cell populations (CB1, Cd45RCd19) and T-cell

populations (CT1, Cd4Cd8). Three B leukemias (B1 and B2,Cd45RCd19; B3, Cd45RlowCd19Sca1) and 3 T leukemias

(T1, Cd4Cd8; T2, Cd4Cd8; and T3, Cd4Cd8) were chosen

for the microarray experiments (National Center for Biotechnology

Information GEO: GSE12581). We were especially interested to

identify genes commonly deregulated in these tumors despite their

heterogeneity and different stage of differentiation. Hierarchical

clustering analyses of genes with a 4-fold change in expression

levels compared with control samples were used to obtain a general

trend (supplemental Figure 1). This analysis allowed us to group

leukemia samples (columns) and/or genes (rows) based on the

similarity of their expression level. As a result, T leukemias and B

leukemias were clustered separately, making 2 distinguishable

groups (supplemental Figure 1). According to these data, Graffi-

induced T and B leukemias showed both distinct and common gene

expression profiles. Indeed, clustering led to the formation of 6

subgroups representing probe sets either specifically deregulated in

B leukemias (188 overexpressed and 86 down-regulated), specifi-

cally deregulated in T leukemias (9 overexpressed and 48 down-

regulated), or commonly deregulated in both types (8 overex-

pressed and 32 down-regulated). The complete list of genes

presenting these lymphoid signatures is available at:

www.biomed.uqam.ca/rassart/microarray2.html.

Gene expression profile specific for pre-B leukemias

We first analyzed the expression profile of genes that could

determine the stage of differentiation of the B leukemias. For allB leukemias, Rag2, Vpreb1, Igll1, Enpep, Ebf1, Il7R, Bst1, and

Foxp1 were overexpressed compared with control cells (Table 1).

Results in this table correspond to the average expression calcu-

lated from all samples analyzed in the microarray analysis (B and

T controls, B lymphoid, T lymphoid, myeloid, erythroid, and

megakaryoblastic tumors). Strong expression of these genes in-

dicated that the 3 Graffi MuLV-induced B tumors were most

probably at a pre-B differentiation stage.5,9-15 In all B leukemias,

we also detected high levels of Cd79a and Cd79b mRNAs and low

levels of Cd20 / Ms4a2 mRNA (Table 1). A total of 274 probe sets,

corresponding to at least 218 genes, were found highly deregulated

in all 3 B leukemias compared with control B lymphocytes. Among

these genes, 72 (86 probe sets) were down-regulated and 146

(188 probe sets) were up-regulated (supplemental Figure 1).

1900 CHARFI et al BLOOD, 10 FEBRUARY 2011 VOLUME 117, NUMBER 6



8/21/2019 Blood-2011-Charfi-1899-910 ppo

4/13

http://bloodjournal.hematologylibrary.org/subscriptions/ToS.dtlhttp://bloodjournal.hematologylibrary.org/

8/21/2019 Blood-2011-Charfi-1899-910 ppo

5/13

Several of them had already been associated with B leukemias and

are listed in Table 2.

To identify new potential oncogenes associated with the devel-

opment of pre-B leukemias, we further focused our analysis on

191 probe sets (supplemental Table 3), mostly because they were

not yet associated with lymphoid leukemias. Some of them

(70 probe sets, including Crisp3 and Lphn2) were already in-

volved in cancer but never associated with B leukemia. Others(72 probe sets) were never associated with cancer, and 36 probe

sets had no assigned function. Finally, the 13 remaining probe sets

were involved in leukemias other than B leukemias (supplemental

Table 3).

Gene expression profile specific for T leukemias

For all analyzed T leukemias, the expression levels of Cd3, Cd4,

and Cd8 markers, compared with the other tumor samples (B lym-

phoid, myeloid, erythroid, and megakaryoblastic tumors), were in

accordance with a T lineage immunophenotype (Table 1). Other

markers known to be present at the T-cell surface were also

detected (Table 1). Among these, Cd6 and Cd28 were the most

specific to Graffi-induced T leukemias. We also observed thatCd160 was expressed in all 3 T leukemias but not in double-

positive control T cells. Moreover, the 2 Cd8 tumors (T1 and T2)

also expressed Cd7 and Dntt (TdT ) (Table 1).

We found that 57 probe sets corresponding to 48 genes showed

the same pattern of deregulation in all 3 T leukemias compared

with control T lymphocytes (supplemental Figure 1; supplemental

Table 3) but not in the B leukemias. Most of them (40 genes) were

down-regulated, whereas only 8 were up-regulated.

Nine probe sets were already associated with T leukemias

(Table 2), thus validating our approach. Among the remaining

48 probe sets, 15 were associated with other cancers, 5 were related

to leukemias other than T leukemias, 20 were neither associated

with leukemias nor with other types of cancer, and 8 had noassigned function (supplemental Table 3).

Expression profiles common to both B and T leukemias

The Robust Multi Array analysis also revealed that 57 probe sets

(corresponding to 48 known genes) were modulated in both T and

B leukemias compared with controls (supplemental Figure 1).

These genes may be associated with common characteristics of

lymphoid leukemias or common oncogenic features and/or directly

related to Graffi leukemogenesis.

Several genes already known to be associated with B and

T leukemias were identified (Table 2). Moreover, among the

57 selected probe sets, 36 had never been simultaneously associ-

ated to both types of lymphoid leukemias although some werealready associated with one type (T leukemia: 10; B leukemia:

4; supplemental Table 3). Among the remaining probe sets, 8 were

identified in other cancers, 9 had never been associated with any

types of cancer, and 4 remain uncharacterized. Matr3 was the only

gene already associated with leukemia (acute myeloid leukemia).

RT-PCR validation

To validate our microarray data, the expression levels of 12 genes

specific to either B, T, or to both leukemias (Fmm2, Arntl2, Bfsp2,

Gfra2, Gmp6a, Gmp6b, Bmp7, Fbln1, Nln, Mettl1, Etv5, and

Celsr1 [Table 3; supplemental Table 3]) were measured by

semiquantitative RT-PCR in several Graffi MuLV-induced tumors

(pure cell populations, Figure 1; or unsorted cell suspensions,

supplemental Figure 2). Significant overexpression of Fmm2,

Arntl2, Bfsp2, and Gfra2 was observed in the majority of B leuke-

mias, although being absent in T leukemias confirming their

specificity to B leukemias (Figure 1A; Table 3; supplemental Table

3). Gmp6a and Gmp6b, expected to be B leukemia-specific,

showed significant overexpression in these tumors compared with

the control samples, albeit in only 2 and 3 of the 5 tested

B leukemias, respectively. As expected, these 2 genes were notexpressed in T leukemias or control samples. High expression

levels of Bmp7 were significantly observed in all T leukemias,

whereas Fbln1 and Nln were overexpressed in 3 of 5 T leukemias

(Figure 1B). Mettl1, expected to be T leukemia-specific gene, was

expressed in both T and B leukemias (including controls; Figure

1B). For Etv5, significant higher levels of expression were ob-

served in 4 of 5 B leukemias and 3 of 5 T leukemias (Figure 1C).

Finally, Celsr1 was significantly overexpressed in most T leuke-

mias compared with normal control (CT2) but, in contradiction

with the microarray data, was poorly expressed in all B leukemias

(Figure 1C). Similar validation results were obtained when using

Cd19 splenic cells as normal control instead of Cd45R cells (not

shown).

Fmn2 induces anchorage-independent growth

We further focused our interest on Fmn2, a gene specifically

overexpressed in B leukemias. This gene is a member of the formin

family (FH proteins), and its product is highly conserved among

the multidomain proteins involved in a growing range of actin-

based processes.16 Most importantly, they may promote cancer

cells to become invasive and metastatic through their actin

remodeling function.17

We first studied the impact of Fmn2 overexpression on the

anchorage-independent growth of NIH/3T3 cells, a classic assay to

demonstrate the oncogenic potential of proteins. As shown in

Figure 2, the number of larger colonies formed in soft agar wassignificantly higher in Fmn2-expressing cells compared with

control cells and similar in cells expressing the human Ras

oncogene (Ras EJ 6.6; Figure 2). These results suggest that the

Fmn2 protein confers anchorage independence to NIH/3T3 cells.

Fmn2 colocalizes with the plasma membrane and disrupts the

actin network

We next determined the subcellular localization of Fmn2. When

GFP-tagged Fmn2 was transiently expressed in NIH/3T3, the

protein was localized at the plasma membrane and within the

cytoplasm (Figure 3A). This was further confirmed by colocaliza-

tion of Fmn2 with a cell membrane marker (CellMask PlasmaMembrane Stains; Figure 3B).

We also observed that NIH/3T3 cells expressing Fmn2 showed

a reduced size and an abnormal morphology compared with control

cells (Figure 3). Moreover, Fmn2-expressing NIH/3T3 cells ap-

peared rounded with long extensions (Figure 3A-D).Actin cytoskel-

eton is one of the cellular components that maintain cell shape and

oncogenic transformation causes profound modifications linked

to cell architecture.18-20 Because the overexpression of Fmn2 in

NIH/3T3 cells altered their morphology, we first tested the effect of

its overexpression on actin filaments and on the microtubule

network because of its reported action on actin cables.21 As

illustrated in Figure 3C-D, cells overexpressing Fmn2 showed a

drastic disruption of the actin and microtubule network and a

reduced number of stress fibers compared with control cells.




8/21/2019 Blood-2011-Charfi-1899-910 ppo

6/13

Table 2. Probe sets already associated with B and T leukemias

Probe set IDs T1-CT* T2-CT* T3-CT* B1-CT† B2-CT† B3-CT† Gene title

Gene

symbol

Down-regulated B leukemia probe sets

1422122_at 0.21 0.06 0.43 6.57 6.89 6.36 Fc receptor, IgE, low affinity II, polypeptide Fcer2a

1451713_a_at 0.05 0.09 0.18 6.26 5.85 5.85 Fc receptor, IgE, low affinity II, polypeptide Fcer2a

1438030_at 0.28 0.01 0.06 5.23 4.94 4.27 RAS, guanyl releasing protein 3 Rasgrp3

1447998_at 0.36 0.14 0.31 4.87 4.53 4.84 Immunoglobulin heavy chain 4 (serum IgG1) Igh-4

1427292_at 0.31 0.37 0.00 4.47 2.90 3.62 Immunoglobulin chain, variable 1 Igl-V1

1439221_s_at 0.18 0.15 0.00 4.44 4.21 4.85 Tumor necrosis factor receptor superfamily, member 5 Tnfrsf5

14 43 783 _x _a t 0 .0 1 0.01 0.16 4.17 4.56 4.67 Histocompatibility 2, class II antigen A, H2-Aa

1452521_a_at 0.01 0.54 0.20 3.98 3.90 4.23 Urokinase plasminogen activator receptor Plaur

1449473_s_at 0.53 0.40 0.27 3.85 3.90 4.69 Tumor necrosis factor receptor superfamily, member 5 Tnfrsf5

1427306_at 0.27 0.08 0.23 3.50 3.45 3.65 Ryanodine receptor 1, skeletal muscle Ryr1

1438031_at 0.41 0.10 0.06 3.49 3.74 3.67 RAS, guanyl releasing protein 3 Rasgrp3

1442263_at 0.15 0.27 0.25 3.48 3.56 3.09 Regulator of G-protein signaling 13 Rgs13

1419609_at 0.32 0.15 0.34 3.45 3.90 3.75 Chemokine (C-C motif) receptor 1 Ccr1

1416055_at 0.04 0.08 0.15 3.43 3.38 3.29 Amylase 2, pancreatic Amy2

1442544_at 0.01 0.06 0.47 3.31 3.53 3.27 Immunoglobulin heavy chain 4 (serum IgG1) Igh-4

1451965_at 0.56 0.42 0.44 3.22 3.69 3.67 Immunoglobulin chain complex Igk

1425063_at 0.07 0.29 0.02 3.15 2.98 2.85 Fc receptor-like 1 Fcrl1

1448861_at 0.50 0.13 0.51 3.15 3.18 3.30 Tnf receptor-associated factor 5 Traf5

1425902_a_at 0.16 0.41 0.48 2.84 2.54 2.90 Nuclear factor of l ight p ol ype pt ide g ene e nh anc er N fk b2

in B cells 2, p49/p100

1425289_a_at 0.10 0.15 0.58 2.82 3.56 4.96 Complement receptor 2 Cr2

1420710_at 0.05 0.37 0.40 2.70 2.12 2.23 Reticuloendotheliosis oncogene Rel

1427576_at 0.52 0.50 0.23 2.60 3.04 3.10 Immunoglobulin chain, constant region Igk-C

1450912_at 0.16 0.23 0.11 2.49 2.19 3.14 Membrane-spanning 4-domains, subfamily A, member 1 Ms4a1

1423226_at 0.03 0.32 0.48 2.32 2.03 2.79 Membrane-spanning 4-domains, subfamily A, member 1 Ms4a1

1455019_x_at 0.30 0.48 0.10 2.43 2.68 2.52 Cytoskeleton-associated protein 4 Ckap4

1421211_a_at 0.38 0.14 0.41 2.37 3.83 3.85 Class II transactivator C2ta

1420353_at 0.36 0.57 0.53 2.36 3.66 3.51 Lymphotoxin A Lta

1422828_at 0.29 0.29 0.43 2.31 2.09 2.21 CD3 antigen, polypeptide Cd3d

1415899_at 0.54 0.24 0.57 2.15 2.50 3.38 Jun-B oncogene Junb

1419714_at 0.42 0.06 0.29 2.11 2.27 2.06 Programmed cell death 1 ligand 1 Pdcd1lg1

1447839_x_at 0.56 0.55 0.08 2.08 2.23 2.22 Adrenomedullin Adm

1425802_a_at 0.26 0.23 0.06 2.07 2.25 4.45 Fc receptor-like A Fcrla

Up-regulated B leukemia probe sets1427276_at 0.20 0.39 0.50 2.03 2.63 2.12 SMC4 structural maintenance of chromosomes 4-like 1 (yeast) Smc4l1

1452499_a_at 0.48 0.09 0.16 2.03 2.17 2.34 Kinesin family member 2A Kif2a

1436036_at 0.40 0.10 0.08 2.08 2.53 2.79 Wolf-Hirschhorn syndrome candidate 1 (human) Whsc1

143 97 40 _s _a t 0. 20 0.01 0.14 2.09 2.62 2.16 Uridine-cytidine kinase 2 Uck2

1448531_at 0.39 0.45 0.36 2.19 2.39 2.27 Lamin B2 Lmnb2

1417299_at 0.24 0.31 0.23 2.38 2.94 2.50 NIMA (never in mitosis gene a)-related expressed kinase 2 Nek2

1420909_at 0.20 0.21 0.01 2.40 3.59 3.57 Vascular endothelial growth factor A Vegfa

1428853_at 0.37 0.09 0.25 2.50 3.08 3.26 Patched homolog 1 Ptch1

1437716_x_at 0.18 0.27 0.08 2.54 2.94 2.43 Kinesin family member 22 Kif22

1418969_at 0.11 0.11 0.05 2.57 3.04 2.53 S-phase kinase-associated protein 2 (p45) Skp2

1415849_s_at 0.30 0.53 0.46 2.61 3.02 2.97 Stathmin 1 Stmn1

1449293_a_at 0.01 0.43 0.22 2.72 2.87 2.65 S-phase kinase-associated protein 2 (p45) Skp2

1423092_at 0.27 0.57 0.49 2.74 3.13 3.10 Inner centromere protein Incenp

1444031_at 0.50 0.17 0.02 2.76 3.05 2.34 Calcium/calmodulin-dependent protein kinase II, Camk2d

1417656_at 0.29 0.06 0.51 2.77 2.90 2.52 Myeloblastosis oncogene-like 2 Mybl21417694_at 0.20 0.39 0. 19 2. 81 2. 83 4 .08 G row th fa ct or r ec ep to r bou nd p rot ei n 2 -a ss oc iat ed p ro te in 1 G ab 1

1439436_x_at 0.01 0.44 0.18 2.83 3.38 3.08 Inner centromere protein Incenp

1424128_x_at 0.30 0.55 0.48 2.84 3.32 3.16 Aurora kinase B Aurkb

1446669_at 0.30 0.45 0.18 2.92 2.95 2.14 Early B-cell factor 1 Ebf1

1440244_at 0.00 0.19 0.05 2.99 4.55 5.28 Avian erythroblastosis virus E-26 (v-ets) oncogene related Erg

1424278_a_at 0.54 0.46 0.47 3.09 3.58 3.10 Baculoviral IAP repeat-containing 5 Birc5

1460247_a_at 0.09 0.12 0.29 3.21 3.47 3.31 S-phase kinase-associated protein 2 (p45) Skp2

1437580_s_at 0.44 0.56 0.10 3.23 3.47 3.29 NIMA (never in mitosis gene a)-related expressed kinase 2 Nek2

1458447_at 0.49 0.43 0.52 3.50 4.00 3.67 Centromere protein F Cenpf

1415810_at 0.19 0.39 0.35 3.50 3.56 3.55 Ubiquitin-like, containing PHD and RING finger domains, 1 Uhrf1

1419943_s_at 0.50 0.52 0.03 3.50 4.23 3.84 Cyclin B1 Ccnb1

Results are presented in log base 2.

T1, T2, and T3 indicateT leukemias; B1, B2, and B3, B leukemias; CT, T control; and CB, B control.

*Ratio T-CT: mean of the deviation ofT1, T2, and T3/T control value.

†Ratio B-CB: mean of the deviation of B1, B2, and B3/B control value.

GENE SIGNATURES OF LYMPHOID LEUKEMIAS 1903BLOOD, 10 FEBRUARY 2011 VOLUME 117, NUMBER 6



8/21/2019 Blood-2011-Charfi-1899-910 ppo

7/13

Table 2. Probe sets already associated with B and T leukemias (continued)


Gene

symbol

1416961_at 0.11 0. 12 0 .34 3. 56 4 .1 4 3. 69 Budd ing un inhi bi te d by b enz imi daz ol es 1 h omo log, Bub1b

(Saccharomyces cerevisiae )

1424991_s_at 0.03 0.04 0.42 3.64 3.98 4.10 Thymidylate synthase Tyms

1417938_at 0.03 0.02 0.37 3.75 4.19 3.88 RAD51-associated protein 1 Rad51ap1

1448899_s_at 0.38 0.31 0.24 3.76 4.13 3.84 RAD51-associated protein 1 Rad51ap1

1416155_at 0.32 0.19 0.26 3.78 4.30 4.38 High mobility group box 3 Hmgb3

1415811_at 0.37 0.03 0.11 3.91 3.95 3.94 Ubiquitin-like, containing PHD and RING finger domains, 1 Uhrf1

1418264_at 0.04 0.39 0.35 4.16 4.52 4.55 SoxLZ/Sox6 leucine zipper binding protein in testis Solt

1417910_at 0.57 0.47 0.32 4.25 4.71 4.28 Cyclin A2 Ccna2

1434748_at 0.27 0.36 0.21 4.30 4.78 4.21 Cytoskeleton-associated protein 2 Ckap2

1448205_at 0.15 0.23 0.15 4.42 4.85 4.49 Cyclin B1, related sequence 1 /// cyclin B1 Ccnb1-rs1///Ccnb1

1416076_at 0.53 0.45 0.19 4.47 5.18 4.81 Cyclin B1 Ccnb1

1452315_at 0.05 0.27 0.57 4.48 4.82 4.58 Kinesin family member 11 Kif11

1439820_at 0.08 0.09 0.13 4.48 4.48 3.95 Early B-cell factor 1 Ebf1

1424046_at 0.04 0.02 0.03 4.56 5.27 4.79 Budding uninhibited by benzimidazoles 1 homolog Bub1


1452314_at 0.14 0.18 0.10 4.58 5.06 4.63 Kinesin family member 11 Kif11

1454694_a_at 0.13 0.08 0.03 4.66 5.14 4.87 Topoisomerase (DNA) II Top2a

1418507_s_at 0.26 0.01 0.06 4.88 5.92 4.41 RIKEN cDNA D130043N08 gene D130043N08Rik

1447363_s_at 0.23 0. 28 0 .1 0 4. 97 5 .5 0 5. 14 Budd in g un in hi bi te d by be nz imi da zol es 1 h omo log, Bub1b


1435306_a_at 0.20 0.32 0.22 5.05 5.50 5.04 Kinesin family member 11 Kif11

1424967_x_at 0.06 0.08 0.27 5.23 3.63 2.48 Troponin T2, cardiac Tnnt2

1418726_a_at 0.29 0.11 0.20 5.49 4.30 3.17 Troponin T2, cardiac Tnnt2

1427161_at 0.31 0.43 0.28 5.67 6.30 5.92 Centromere autoantigen F Cenpf

1426817_at 0.25 0.39 0.41 5.64 6.09 5.88 Antigen identified by monoclonal antibody Ki 67 Mki67

1420176_x_at 0.27 0.14 0.04 7.08 7.05 7.00 Immunoglobulin -like polypeptide 1 Igll1

1448649_at 0.02 0.24 0.17 9.58 9.56 9.18 Glutamyl aminopeptidase Enpep

Down-regulated T leukemia probe sets

1427419_x_at 2.15 8.36 8.79 0.26 0.06 0.26 Chemokine (C-C motif) receptor 9 Ccr9

1421919_a_at 2.68 6.65 6.65 0.03 0.06 0.30 Chemokine (C-C motif) receptor 9 Ccr9

1425792_a_at 2.79 3.33 5.75 0.18 0.26 0.30 RAR-related orphan receptor gamma Rorc

1423954_at 3.21 2.52 3.89 0.26 0.13 0.24 Complement component 3 C3

1447541_s_at 2.12 2.46 2.49 0.53 0.33 0.30 Integrin, E, epithelial-associated Itgae

1450295_s_at 2.23 2.13 2 .8 8 0. 11 0 .3 6 0 .13 Poli ov irus r ec ep tor // / D NA se gme nt , C hr 7, ERATO D oi Pvr // / D 7Ertd 45 8e458, expressed

1427411_s_at 2.60 2.07 2.24 0.33 0.31 0.06 Deleted in lymphocytic leukemia, 2 Dleu2

1420413_at 2.30 2.07 2 .0 6 0. 42 0. 29 0 .50 Aolu te c ar ri er f ami ly 7 ( ca ti oni c a mi no ac id t ran spor ter, Slc 7a 11

y system), member 11

Up-regulated T leukemia probe sets

1449835_at 2.55 2.86 5.05 0.30 0.38 0.35 Programmed cell death 1 Pdcd1

Down-regulated B and T leukemia common probe sets

1427127_x_at 6.03 4.87 2.83 5.10 5.83 5.70 Heat shock protein 1B /// heat shock protein 1A Hspa1b /// Hspa1a

1452318_a_at 5.74 4.92 3.09 4.73 5.26 5.81 Heat shock protein 1B Hspa1b

1427126_at 5.23 4.56 2.76 4.92 5.53 5.75 Heat shock protein 1B Hspa1b

1448123_s_at 3.86 3.54 3.37 2.85 2.70 2.36 Transforming growth factor, induced Tgfbi

1448162_at 3.84 3.92 3.30 4.15 3.86 4.01 Vascular cell adhesion molecule 1 Vcam1

1456250_x_at 3.57 3.18 3.79 2.29 2.18 2.24 Transforming growth factor, induced Tgfbi

1460423_x_at 3.39 2.89 2.13 6.19 6.45 7.35 Ig chain IgM

1418126_at 2.89 2.92 2.91 2.02 2.34 2.80 Chemokine (C-C motif) ligand 5 Ccl51417756_a_at 2.84 5.14 3.04 2.50 3.17 2.19 Lymphocyte specific 1 Lsp1

1449858_at 2.83 2.82 2.50 4.78 4.68 2.89 CD86 antigen Cd86

1449164_at 2.76 3.00 3.03 2.83 2.94 2.67 CD68 antigen Cd68

1418365_at 2.71 3.02 2.34 4.61 4.43 5.66 Cathepsin H Ctsh

1415989_at 2.46 2.50 2.69 2.07 2.32 2.37 Vascular cell adhesion molecule 1 Vcam1

1434499_a_at 2.44 2.70 6.15 2.72 2.91 2.62 Lactate dehydrogenase 2, B chain Ldh2

1448237_x_at 2.43 2.82 5.89 2.01 2.07 2.07 Lactate dehydrogenase 2, B chain Ldh2

1450381_a_at 2.23 3.10 2.91 2.74 3.38 3.19 B-cell leukemia/lymphoma 6 Bcl6

1450381_a_at 2.23 3.10 2.91 2.74 3.38 3.19 B-cell leukemia/lymphoma 6 Bcl6


T1, T2, andT3 indicate T leukemias; B1, B2, and B3, B leukemias; CT, T control; and CB, B control.

*Ratio T-CT: mean of the deviation of T1, T2, andT3/T control value.





8/21/2019 Blood-2011-Charfi-1899-910 ppo

8/13

FMN2 gene overexpression in human pre-B leukemia

We measured the expression levels of human FMN2 in patients

with different types of B leukemias (Figure 4). No expression was

detected by RT-PCR in control samples, in mantle cell lymphoma,

follicular lymphoma, B-cell prolymphocytic leukemia, and chronic

lymphocytic leukemia. In contrast, FMN2 expression was easily

detected in almost all pre-B-ALL patients (L1 and L2; 18 of 19)and in one of 2 Burkitt leukemias. Interestingly, the highest levels

of FMN2 were detected in all L1 pediatric pre-B-ALL samples

bearing a t(12;21) translocation (lanes 7, 9, 11, and 12). Another

tumor (lane 3) also showed high levels of FMN2 expression,

suggesting that it could harbor the translocation as well, although

this remained to be confirmed.

Discussion

Graffi MuLV is a good model to gain new insights on leukemia

development and progression and to identify new oncogenes. Gene

expression profiling of each type of leukemia (T lymphoid,

B lymphoid, myeloid, erythroid, and megakaryoblastic) served toidentify the cancer signatures of these specific leukemias.5 In

this paper, we determined the expression profile of 3 B and

3 T leukemias induced by this retrovirus compared with nonleuke-

mic cell controls (CB and CT, Tables 2, 3; supplemental Table 3)

and to nonlymphoid leukemias induced by the same virus (Table 1;

www.biomed.uqam.ca/rassart/microarray2.html). Setting the mini-

mal acceptable change in expression levels to 4-fold, we selected

388 probe sets corresponding to 305 genes: 48 genes specifically

modulated in T leukemias, 218 in B leukemias, and 40 in both types

compared with their respective controls.

Phenotypic properties and cancer signature of B leukemias

Our analysis suggests that B leukemias induced by Graffi MuLV

are arrested at the pre-B stage of differentiation based on the high

expression of pre-B specific markers (Table 1). As in human

pre-B-ALL, these leukemias overexpressed surface markers, such

as Cd79a and Cd79b, and lacked Cd20 / Ms4a222 (Table 1). These

results suggest that data retrieved from the gene profiling analy-

sis of B leukemias should be especially relevant for human

pre-B-ALL. Indeed, the results obtained for FMN2 expression in

human leukemias are in total agreement with the mouse data.

We selected 218 genes that were specifically deregulated in all

3 B leukemias; and as presented in Table 2, several of these genes

were already known to be involved in lymphoid leukemia. This not

only validates our microarray analysis but also further defines the

cancer signature of these B lymphoid leukemias.

Phenotypic properties and cancer signature of T leukemia

Three distinct T leukemias (T1, Cd4Cd8; T2, Cd4Cd8; and T3,

Cd4Cd8) were chosen for the microarray experiments. These

3 different phenotypes are frequently observed in Graffi-induced

Table 3. Genes selected for RT-PCR validation

Probe set IDs T1-CT* T2-CT* T3-CT* B1-CT* B2-CT* B3-CT* Gene title

Gene

symbol

B leukemia probe sets

1450063_at 0.33 0.20 0.19 2.59 3.05 3.01 Formin 2 Fmn2

1429688_at 0.48 0.13 0.38 3.24 2.78 2.75 Aryl hydrocarbon receptor nuclear translocator-like 2 Arntl2

1434463_at 0.45 0.48 0.29 2.98 3.36 3.39 Beaded filament structural protein 2, phakinin Bfsp21423007_a_at 0.41 0.04 0.01 3.12 3.22 2.68 Glial cell line derived neurotrophic factor family receptor Gfra2

1426442_at 0.06 0.04 0.09 3.15 2.79 2.42 Glycoprotein m6a Gpm6a

1425942_a_at 0.43 0.54 0.09 6.93 6.35 2.55 Glycoprotein m6b Gpm6b

T leukemia probe sets

1419122_at 2.33 2.16 2.85 0.41 0.07 0.15 Methyltransferase-like 1 Mettl1

1451555_at 2.32 3.23 2.07 0.06 0.07 0.09 Neurolysin (metallopeptidase M3 family) Nln

1432410_a_at 3.09 1.56 3.28 0.36 0.15 0.19 Bone morphogenetic protein 7 Bmp7

1451119_a_at 2.75 2.25 1.58 0.12 0.28 0.27 Fibulin 1 Fbln1

B and T leukemia common probe sets

1428142_at 2.68 2.24 3.42 4.04 3.53 3.66 ets variant gene 5 Etv5

1418925_at 2.38 1.50 1.20 2.99 2.99 2.41 Cadherin EGF LAG seven-pass G-type receptor 1 Celsr1

Results are presented in log base 2. A positive deviation of 4 and above and a negative deviation of 4 and below indicates a fold change of 4. Values between 0.585 and

0.585 are considered to be not significant.

T1, T2, and T3 indicate T leukemias; B1, B2, and B3, B leukemias; CT, T control; and CB, B control.

*Ratio T-CTand B-CB: mean of the deviation of T1, T2, and T3/T control value and mean of the deviation of B1, B2, and B3/B control value.

Table 2. Probe sets already associated with B and T leukemias (continued)


Gene

symbol

Up-regulated B and T leukemia common probe sets

1422967_a_at 2.03 2.96 2.87 2.32 3.45 2.72 Transferrin receptor Tfrc

1422198_a_at 2.03 2.21 2.54 2.39 3.05 2.17 Serine hydroxymethyl transferase 1 (soluble) Shmt1

1425179_at 2.15 2.31 2.71 2.80 3.32 2.52 Serine hydroxymethyl transferase 1 (soluble) Shmt1

1425923_at 6.04 5.92 5.57 4.66 4.87 4.19 v-myc myelocytomatosis viral-related oncogene, Mycn

neuroblastoma derived (avian)


T1, T2, and T3 indicate T leukemias; B1, B2, and B3, B leukemias; CT, T control; and CB, B control.

*Ratio T-CT: mean of the deviation of T1, T2, and T3/T control value.





8/21/2019 Blood-2011-Charfi-1899-910 ppo

9/13

T leukemias and are also representative of normal T-cell types in

healthy mice.1 The retrovirus may have targeted slightly different

T-cell blasts or transformed the targeted cells through these

slightly different lineages. Based on the analysis, all T leukemias

expressed markers specific for the T lineage (Table 1) as well as

T-ALL–associated markers, such as Cd7 and Dntt (TdT ).23 How-

ever, we also found that all 3 T leukemias expressed high levels of

Cd160. This marker, normally expressed on human NK and only on

a subtype of human CD8 cells,24 seems to be overexpressed in

almost all cases of B-cell chronic lymphocytic leukemia.25

The analysis reveals that a total of 48 genes were strictly

modulated in all T leukemias compared with control T cells (Tables

2-3; supplemental Table 3), including 9 probe sets already associated

with leukemias. The overall data certainly reveals the existence of a

Figure 1. Analysis of selected genes differentially expressed in sorted lymphoid leukemia samples. Semiquantitative RT-PCR analysis in 5 T (T4, Cd4 Cd8 ; T5,

Cd4 Cd8 ;T6, Cd4 Cd8 ;T7, Cd4 Cd8 ; T8, Cd4 Cd8 )and5 B (B4, Cd45 Cd19 Sca1; B5, Cd45R Cd19 Sca1; B6, Cd45R Cd19 Sca1; B7, Cd45R Cd19 Sca1;

B8, Cd45R Cd19 Sca1) leukemias. RT-PCRs were performed in triplicate for each gene. The actin gene was used as internal control in specific conditions described in

“Semiquantitative RT-PCR” and expression level in each leukemia is presented as a selected gene/actin density ratio. (A) B leukemia-specific genes. (B) T leukemia-specific

genes. (C) Genes common to both leukemias. Statistical analysis was performed using one-way analysis of variance, and P less than .05 was considered to be significant

(*P .05, **P .01, ***P .001) compared with the respective control (B cells from normal spleen [CB2] for the B leukemias and T cells from normal thymus [CT2] for the

T leukemias).




8/21/2019 Blood-2011-Charfi-1899-910 ppo

10/13

common leukemic gene signature between T leukemias induced by

the virus despite heterogeneity and phenotypic differences. We can

speculate that similar deregulation patterns observed in these

leukemias are associated with common characteristics of T leuke-

mias or common oncogenic properties and are probably found in

human T leukemias as well.

Cancer signature common to both B and T leukemias

Because we simultaneously analyzed gene expression profiles of

B and T leukemias induced by Graffi MuLV, regardless of lineage

differences and tumor heterogeneity, we found that 57 others probe

sets modulated in T leukemias were similarly modulated in

B leukemias, including 21 previously reported genes (Table 2).

These results strongly reinforce the potential of the remaining

36 probe sets to be important for the development of B leukemias.

Among these, 23 probe sets had already been associated with either

T or B leukemias, with myeloid leukemia or other cancers

(supplemental Table 3). These genes are certainly part of the

common cancer signature of the T and B lymphoid leukemias and

may be relevant for human lymphoid leukemias.

New potential markers and candidate oncogenes from

B and T leukemias

The combination of our Graffi-induced tumor model and DNA

microarrays allowed us to identify potential new markers of B and

T leukemias that could also play an oncogenic role. Among theselected 388 probe sets (at least 305 genes), we further identified

275 genes not yet associated with lymphoid leukemias (supplemen-

tal Table 3). Some were specific to B or T leukemias or common to

both types, and 124 were obvious oncogene candidates because

they had been already associated with nonlymphoid leukemia or

other cancers. More interestingly, we identified a total of 103 genes

that had never been associated with any type of cancer and 32 probe

sets corresponding to uncharacterized genes or genes with un-

known function.

By RT-PCR, we have validated changes in the expression levels

of 10 of the 12 genes selected according to their specificity for

lymphoid leukemias (Figure 1; supplemental Figure 2). Our

analysis revealed deregulation of the expression of several genes

associated with Graffi MuLV-induced B leukemias (Table 1; Figure

1A; supplemental Figure 2). These genes are Fmm2, Arntl2 (from

the bHLH-PAS superfamily involved in regulating cell growth and

differentiation26), Bfsp2 (a member of the intermediate filament

family and component of cytoskeleton proteins in the lens cells,

maintaining their morphology and promoting their motility27),

Gfra2 (from the glial cell line-derived neurotrophic factor recep-

tor- family28 and associated with primary neuroblastomas29 and

with some medullary thyroid carcinoma tumor cells30), and Gmp6a

and Gmp6b (members of the myelin proteolipid protein family and

neuronal homologues of PLP/DM2031).

We also identified several genes that were associated with

T leukemias (Figure 1B), namely, Nln (from the metallopeptidase

M3 family and involved in the metabolism of neurotensin32), Bmp7

(a cytokine from the transforming growth factor- superfamilyand expressed in various types of cancer, including prostate and

breast cancers and melanoma33-35), and Fbln1 (involved in heart

development and in cell signaling involving growth factors36 and

associated with human neoplasia, especially breast and ovarian

cancers37).

Etv5 (a member of the Ets family of transcription factors) was

already described in association with B leukemias38 but, according

to our data, it appears also overexpressed in T leukemias (Figure

1C). In contrast, the modulation of expression of Celsr1 (involved

in the regulation of cell polarity and in convergent extension39 and

expressed in gastrointestinal tumors40), expected to be specific to

both B and T leukemias based on the microarray analysis, was

validated in all T leukemic samples (Figure 1C; supplementalFigure 2) but in only one B tumor (supplemental Figure 2).

Finally, the high level of expression of Mettl1 (highly expressed

in lung cancer41) observed by the microarray analysis was not

confirmed by RT-PCR (Table 3; supplemental Table 3; Figure 1B)

and, as for Celsr1, may reflect a certain degree of tumor heteroge-

neity as often observed in human leukemias as well.

We also compared gene expression between B tumors and

normal B cells from the bone marrow sorted with an anti-Cd45R

antibody. This control includes B cells at different stages of

maturation. The majority of the tested genes showed results similar

to those obtained with spleen the B cells as control except for

Fmn2, which was highly expressed in the bone marrow-derived

cells (data not shown). Although this had not yet been reported,

Fmn2 seems to be normally expressed at an early stage of B-cell

Figure 2. Effect of Fmn2protein expression on anchorage of NIH/3T3

cells. Cells were transiently transfected with pCMV empty vector, with the

pCMV-Fmn2 vector, and the Ras (EJ 6.6) expression vector. Transfected

or control cells (104) were plated in soft agar as described in “Colony

formation in soft agar.” After 4 weeks, the number of colonies (at least

twice larger than colonies from controls) was scored (A). The results

represent the average of 3 independent experiments. Statistical analysis

was performed using one-way analysis of variance (P .05). Cells were

observed with an optical microscope (Ernst Leitz, 6MBH Wetzlar), and

representative fields were photographed using a numerical camera

(Nikon coolpix 4500; original magnification 40). Images were analyzed

using NIH ImageJ software Version 1.42l. Cells were left untransfected

(B) or transfected with Ras EJ 6.6 (C), empty vector (D), or pCMV-Fmn2

(E).




8/21/2019 Blood-2011-Charfi-1899-910 ppo

11/13

lymphopoiesis in the bone marrow. Its expression decreases when

B cells move from the bone marrow to the spleen for maturation.

This inverse correlation between the expression level of Fmn2

and B-cell maturation does not necessarily contradict its involve-

ment in carcinogenesis. This is exemplified with GATA2, which

is essential for the maintenance and the proliferation of hematopoi-

etic progenitors during normal hematopoiesis42 while having been

implicated in tumorigenesis.

Overall, these results strongly suggest that the majority of the

275 selected probe sets, in particular Fmn2, Arntl2, Gpm6a,

Gpm6b, Bfsp2, Gfra2, Nln, Bmp7, Fbln1, and Etv5, are potentially

new specific markers or oncogenes for B, T, or B and T leukemias.

Our analysis also identified several down-regulated genes, such as

Klk6 and Tgf I (Table 2), which are already identified as tumors

suppressors. Further studies are required to determine whether the

modulation of expression also correlates with changes at the

protein level and most importantly to determine their potential role

in human leukemias. Regarding Fmn2, our attempts to measure

levels of this protein in mice tumors was hampered by the lack of

specific antibodies.

Fmn2 gene is a good candidate oncogene

Among the 10 genes validated by RT-PCR, we further character-

ized Fmn2, which was specifically associated with B leukemias.

Fmn2 is expressed in the developing and mature central nervous

system43 and in oocytes.16 It was identified as a formin homology

(FH) gene and the protein contains 2 characteristic FH proteindomains: FH1 (proline-rich region) and FH2. The latter is respon-

sible for actin nucleation.44 The comparison of the mouse and

human Fmn2 showed 74.7% sequence homology. Members of the

formin family are implicated in cytokinesis, organogenesis, and

normal tissue homeostasis but are also involved in the invasive

potential of cancerous cells and metastasis.17 The implication of

Fmn2 in the development of tumors had not yet been demonstrated,

even though human FMN2 ESTs were found in several human

tumors (parathyroid tumor, glioblastoma, retinoblastoma, and

chondrosarcoma).17

In this paper, we report, for the first time, that Fmn2 is not only

specifically overexpressed in B leukemias induced by the Graffi

virus in mice (Figure 1; Table 3; supplemental Table 3) but more

importantly in human pre-B-ALL (Figure 4).

Figure 3. Subcellular localization of Fmn2 and its

effect on cytoskeleton. The GFP-tagged Fmn2 protein

was localized in NIH/3T3 cells and (A) tested for its

effect shown on the shape of the cells. (B) Plasma

membrane labeling with CellMask. (C) Actin labeling

with AlexaFluor-555–conjugated phalloidin. (D) -

tubulin labeling with anti–-tubulin antibody. Images were

captured by a laser-scanning confocal microscope (Bio-

RadMRC-1024 ES)mountedon a NikonTE-300usinga

60 /1.4 NA oil Plan Apo VC objective, digitally acquired

using Laser Sharp software Version 3.2 (Bio-Rad), and

analyzedusing NIHImageJ Version1.42l software. Data

are representative of 3 independent experiments. The

GFP vector alone was used as a control. Ovals and

arrows indicate transfected and nontransfected cells,

respectively.




8/21/2019 Blood-2011-Charfi-1899-910 ppo

12/13

Moreover, we demonstrate that ectopic expression of Fmn2

confers anchorage-independent growth to NIH3T3 cells (Figure 2).

This anchorage-independent growth conferred by Fmn2 is prob-

ably related to its ability to induce the disruption of the actin and

microtubule network and a reduction of the number of stress fibers.

The biologic role of Fmn2 in actin and microtubule network

disruption in B leukemia is not quite clear. However, we are

convinced that the up-regulation of FMN2 expression could disturb

the dynamic of the actin network of ALL cells accompanied by the

reorganization of their cytoskeleton, which in turn could contribute

to their abnormal behaviors. Some examples highlight the fact that

even B cells change form depending on their state of developmentor their abnormal behaviors. Indeed, it was reported that during

spreading, the B lymphocyte cytoarchitecture is converted from a

semirigid (before migration) to a more flexible state (during

migration). This migration seems to involve up-regulation of CD44

adhesion molecule on activated B lymphocytes and to require the

rearrangement of many cytoskeleton components (actin, microtu-

bules, and vimentin).45 In addition, Caligaris-Cappio et al demon-

strated that cells from B-chronic lymphocytic leukemia or hairy

cell leukemia (HCL) showed an aberrant cytoskeleton organiza-

tion.46 In addition, Schmitt-Gräff et al observed changes in the

F-actin in B cells of patients with ALL.47

Implication of FMN2 in human pre-B-ALL

To determine the possible contribution of human FMN2 to

leukemogenesis, we measured its expression in 25 different

B leukemia samples. We showed that this gene was specifically

overexpressed in L1 and L2 pre-B-ALL (18 of 19 of cases; Figure

4), thereby agreeing with our microarray data (Table 1). More

importantly, we show that 4 pediatric pre-B-ALL samples with a

t(12;21) translocation produced the strongest signals. Such t(12;21)

rearrangement involving the TEL / AML1 genes is more frequent in

childhood ALL (25%–30%) with a B-precursor phenotype than in

adult ALL (3%–5%).48 Although this translocation is associated

with a favorable outcome and a good response to conventional

chemotherapy, 25% of relapses occur off-therapy and require

additional therapeutic strategies. The strong expression of FMN2 in

pediatric pre-B-ALL with the t(12;21) translocation suggests thatthe TEL/AML1 fusion protein could directly or indirectly up-

regulate FMN2 gene expression. Together, these results suggest

that very high expression of FMN2 in pre-B-ALL could be

correlated with a t(12;21) translocation and could be used as marker,

although this has to be confirmed with a larger panel of samples.

In conclusion, we identified a set of genes that are specific

markers for B and T leukemias induced by the Graffi MuLV and

thus may also serve as potential markers in human lymphoid

leukemias. Some of these genes may have oncogenic properties as

revealed with the mouse Fmn2 gene. For the first time, we show

that FMN2 is up-regulated in human pre-B-ALL and more specifi-

cally in pediatric pre-B-ALL with the t(12;21) translocation.

Additional investigations are necessary to further characterize its

function in tumor induction.

Acknowledgments

The authors thank Dr Daniel Sinnet for providing pediatric tumor

samples; André Ponton, Michal Blazejczyk, and Mathieu Miron

from Genome Quebec Innovation Center (Montreal, QC) for help

with the design and analyses of the microarray experiments; Dr

Benoit Barbeau for critical reading of the manuscript; and Denis

Flipo for help with the confocal microscopy analysis.

This work was supported by Canadian Institutes of Health

Research (grant MOP-37994; E.R.). C.C. is a recipient of a

studentship from the Tunisia Government and Fondation UQAM.

Authorship

Contribution: E.R. and V.V. designed the microarray experiments;

V.V. performed the microarray experiments; C.C. analyzed the

microarray data of the lymphoid leukemias and performed the

experiments; L.-C.L. contributed to some experiments and helpful

discussions; C.C., E.E., and E.R. wrote the manuscript; and E.E

and E.R. supervised the overall project.

Conflict-of-interest disclosure: The authors declare no compet-

ing financial interests.

Correspondence: Eric Rassart, Département des Sciences Bio-logiques, Université du Québec à Montréal, Case Postale 8888

Succursale Centre-ville, Montréal, QC, H3C-3P8, Canada; e-mail:

[email protected]; and Elsy Edouard, Département des Sci-

ences Biologiques, Université du Québec à Montréal, Case Postale

8888 Succursale Centre-ville, Montréal, QC, H3C-3P8, Canada;

e-mail: [email protected].

References

1. Voisin V, Barat C, Hoang T, Rassart E. Novel in-

sights into the pathogenesis of the Graffi murine

leukemia retrovirus. J Virol. 2006;80(8):4026-4037.

2. Denicourt C, Edouard E, Rassart E. Oncogene

activation in myeloid leukemias by Graffi murine

leukemia virus proviral integration. J Virol. 1999;

73(5):4439-4442.

3. Denicourt C, Kozak CA, Rassart E. Gris1, a new

common integration site in Graffi murine leukemia

virus-induced leukemias: overexpression of a

truncated cyclin D2 due to alternative splicing.

J Virol. 2003;77(1):37-44.

4. Denicourt C, Legault P, McNabb FA, Rassart E.

Human and mouse cyclin D2 splice variants:

transforming activity and subcellular localization.

Oncogene. 2008;27(9):1253-1262.

5. Voisin V, Legault P, Salazar Ospina DP, Ben-

David Y, Rassart E. Gene profiling of the erythro-

and megakaryoblastic leukaemias induced by the

Graffi murine retrovirus. BMC Med Genomics.

2010;3(1):2.

Figure 4. Expression analysis of human FMN2 gene in different types of B

leukemic samples. Semiquantitative RT-PCR analysis was performed on 12 pediatric

pre-B-ALL (lanes 1-12), 7 adult pre-B-ALL (lanes 13-19), 2 Burkitt leukemias (lanes

20 and 21), one mantle cell lymphoma (lane 22), one follicular lymphoma (lane 23),

one B-cell prolymphocytic leukemia (lane 24), and one chronic lymphocytic leukemia

(lane 25). RT-PCRs were performed in triplicate for each gene. The actin gene

was used as an internal control as described in “Semiquantitative RT-PCR,” and

expression levelin eachleukemiais presentedas a selectedgene/actin density ratio.

Statistical analysis was performed using one-way analysis of variance, and P less

than .05 was considered to be significant (*P .05, **P .01, ***P .001) com-

pared with the respective control (CH).




8/21/2019 Blood-2011-Charfi-1899-910 ppo

13/13

6. Blazejczyk M, Miron M, Nadon R. FlexArray: a

statistical data analysis software for gene expres-

sion microarrays. Genome Quebec, Montreal,

Canada. http://genomequebec.mcgill.ca/

FlexArray.

7. Eisen MB, Spellman PT, Brown PO, Botstein D.

Cluster analysis and display of genome-wide ex-

pression patterns. Proc Natl Acad Sci U S A.

1998;95(25):14863-14868.

8. Landais S, Landry S, Legault P, Rassart E. Onco-

genic potential of the miR-106–363 cluster and its

implication in human T-cell leukemia. Cancer

Res. 2007;67(12):5699-5707.

9. Bories JC, Cayuela JM, Loiseau P, Sigaux F.

Expression of human recombination activating

genes (RAG1 and RAG2) in neoplastic lymphoid

cells: correlation with cell differentiation and anti-

gen receptor expression. Blood. 1991;78(8):

2053-2061.

10. Donohoe ME, Blomberg BB. The 14.1 surrogate

light chain promoter has lineage- and stage-

restricted activity. J Immunol. 1997;158(4):1681-

1691.

11. Hirata Y, Kimura N, Sato K, et al.ADP ribosyl cy-

clase activity of a novel bone marrow stromal cell

surface molecule, BST-1. FEBS Lett. 1994;

356(2):244-248.

12. Hu H, Wang B, Borde M, et al. Foxp1 is an essen-tial transcriptional regulator of B celldevelopment.

Nat Immunol. 2006;7(8):819-826.

13. Li L, Wang J, Cooper MD. cDNA cloning and ex-

pression of human glutamyl aminopeptidase

(aminopeptidase A). Genomics. 1993;17(3):657-

664.

14. Ohnishi K, Takemori T. Molecular components

and assembly of mu.surrogate light chain com-

plexes in pre-B cell lines. J Biol Chem. 1994;

269(45):28347-28353.

15. Thal MA, Carvalho TL, He T, et al. Ebf1-mediated

down-regulation of Id2 and Id3 is essential for

specification of the B cell lineage. Proc NatlAcad

Sci U S A. 2009;106(2):552-557.

16. Faix J, Grosse R. Staying in shape with formins.

Dev Cell. 2006;10(6):693-706.

17. Katoh M, Katoh M. Characterization of FMN2gene at human chromosome 1q43. Int J Mol Med.

2004;14(3):469-474.

18. Amsellem V, Kryszke MH, Hervy M, et al. The

actin cytoskeleton-associated protein zyxin acts

as a tumor suppressor in Ewing tumor cells. Exp

Cell Res. 2005;304(2):443-456.

19. Drubin DG, Nelson WJ. Origins of cell polarity.

Cell. 1996;84(3):335-344.

20. Pawlak G, Helfman DM. Cytoskeletal changes in

cell transformation and tumorigenesis. Curr Opin

Genet Dev. 2001;11(1):41-47.

21. Wallar BJ,Alberts AS. The formins: active scaf-

folds that remodel the cytoskeleton. Trends Cell

Biol. 2003;13(8):435-446.

22. Chu PG,Arber DA. CD79: a review. Appl Immu-

nohistochem Mol Morphol. 2001;9(2):97-106.

23. LangerakAW, Wolvers-Tettero IL, van den

Beemd MW, et al. Immunophenotypic and immu-

nogenotypic characteristics of TCRgammadelta

T cell acute lymphoblastic leukemia. Leukemia.

1999;13(2):206-214.

24. Agrawal S, Marquet J, Freeman GJ, et al. Cutting

edge: MHC class I triggering by a novel cell sur-

face ligand costimulates proliferation of activated

human T cells. J Immunol. 1999;162(3):1223-

1226.

25. Liu FT, Giustiniani J, Farren T, et al. CD160 sig-

naling mediates PI3K-dependent survival and

growth signals in chronic lymphocytic leukemia.

Blood. 2010;115(15):3079-3088.

26. Yeh CT, Lu SC, Tseng IC, et al. Antisense overex-

pression of BMAL2 enhances cell proliferation.

Oncogene. 2003;22(34):5306-5314.

27. Ma X, Li FF, Wang SZ, Gao C, Zhang M, Zhu SQ.

A new mutation in BFSP2 (G1091A) causes auto-

somal dominant congenital lamellar cataracts.

Mol Vis. 2008;14:1906-1911.

28. Quartu M, Serra MP, Boi M, Ferretti MT, Lai ML,

Del Fiacco M. Tissue distribution of Ret,

GFRalpha-1, GFRalpha-2 and GFRalpha-3 re-

ceptors in the human brainstem at fetal, neonatal

and adult age. Brain Res. 2007;1173:36-52.

29. Hishiki T, Nimura Y, Isogai E, et al. Glial cell line-

derived neurotrophic factor/neurturin-induced dif-ferentiation and its enhancement by retinoic acid

in primary human neuroblastomas expressing

c-Ret, GFR alpha-1, and GFR alpha-2. Cancer

Res. 1998;58(10):2158-2165.

30. Lindahl M, Poteryaev D, Yu L, et al. Human glial

cell line-derived neurotrophic factor receptor al-

pha 4 is the receptor for persephin and is pre-

dominantly expressed in normal and malignant

thyroid medullary cells. J Biol Chem. 2001;276(12):

9344-9351.

31. Werner H, Dimou L, Klugmann M, Pfeiffer S,

Nave KA. Multiple splice isoforms of proteolipid

M6B in neurons and oligodendrocytes. Mol Cell

Neurosci. 2001;18(6):593-605.

32. Brown CK, Madauss K, Lian W, Beck MR, Tolbert

WD, Rodgers DW. Structure of neurolysin reveals

a deep channel that limits substrate access. Proc

NatlAcad Sci U S A. 2001;98(6):3127-3132.33. Alarmo EL, Kuukasjarvi T, Karhu R, Kallioniemi A.

A comprehensive expression survey of bone mor-

phogenetic proteins in breast cancer highlights

the importance of BMP4 and BMP7. Breast Can-

cer Res Treat. 2007;103(2):239-246.

34. Doak SH, Jenkins SA, Hurle RA, et al. Bone mor-

phogenic factor gene dosage abnormalities in

prostatic intraepithelial neoplasia and prostate

cancer. Cancer Genet Cytogenet. 2007;176(2):

161-165.

35. Rothhammer T, Poser I, Soncin F, Bataille F,

Moser M, BosserhoffAK. Bone morphogenic pro-

teins are overexpressed in malignant melanoma

and promote cell invasion and migration. Cancer

Res. 2005;65(2):448-456.

36. Perbal B, Martinerie C, Sainson R, Werner M, He

B, Roizman B. The C-terminal domain of the

regulatory protein NOVH is sufficient to promote

interaction with fibulin 1C: a clue for a role of

NOVH in cell-adhesion signaling. Proc NatlAcad

Sci U S A. 1999;96(3):869-874.

37. Greene LM, Twal WO, Duffy MJ, et al. Elevated

expression and altered processing of fibulin-1

protein in human breast cancer. Br J Cancer.

2003;88(6):871-878.

38. Korz C, Pscherer A, Benner A, et al. Evidence fordistinct pathomechanisms in B-cell chronic lym-

phocytic leukemia and mantle cell lymphoma by

quantitative expression analysis of cell cycle and

apoptosis-associated genes. Blood. 2002;99(12):

4554-4561.

39. Katoh Y, Katoh M. Comparative integromics on

FAT1, FAT2, FAT3 and FAT4. Int J Mol Med.

2006;18(3):523-528.

40. Katoh M, Katoh M. Comparative integromics on

non-canonical WNT or planar cell polarity signal-

ing molecules: transcriptional mechanism of

PTK7 in colorectal cancer and that of SEMA6A in

undifferentiated ES cells. Int J Mol Med. 2007;

20(3):405-409.

41. Wikman H, Nymark P, Vayrynen A, et al. CDK4 is

a probable target gene in a novel amplicon at

12q13.3-q14.1 in lung cancer. Genes Chromo-

somes Cancer. 2005;42(2):193-199.

42. Ferreira R, Ohneda K, Yamamoto M, Philipsen S.

GATA1 function, a paradigm for transcription fac-

tors in hematopoiesis. Mol Cell Biol. 2005;25(4):

1215-1227.

43. Leader B, Leder P. Formin-2, a novel formin ho-

mology protein of the cappuccino subfamily, is

highly expressed in the developing and adult cen-

tral nervous system. Mech Dev. 2000;93(1):221-

231.

44. Dumont J, Million K, Sunderland K, et al.

Formin-2 is required for spindle migration and for

the late steps of cytokinesis in mouse oocytes.

Dev Biol. 2007;301(1):254-265.

45. Sumoza-Toledo A, Santos-Argumedo L. The

spreading of B lymphocytes induced by CD44

cross-linking requires actin, tubulin, and vimentinrearrangements. J Leukoc Biol. 2004;75(2):233-

239.

46. Caligaris-Cappio F, Bergui L, Tesio L, Corbascio

G, Tousco F, Marchisio PC. Cytoskeleton organi-

zation is aberrantly rearranged in the cells of B

chronic lymphocytic leukemia and hairy cell leu-

kemia. Blood. 1986;67(1):233-239.

47. Schmitt-Gräff A, Chaponnier C, Gabbiani G. Cy-

toskeletal organization of peripheral blood normal

and leukemic lymphocytes and lymphoblasts.

J Submicrosc Cytol. 1987;19(2):329-335.

48. De Braekeleer E, Basinko A, Douet-Guilbert N, et

al. Cytogenetics in pre-B and B-cell acute lym-

phoblastic leukemia: a study of 208 patients diag-

nosed between 1981 and 2008. Cancer Genet

Cytogenet. 2010;200(1):8-15.



blood-2011-charfi-1899-910 ppo

Documents