E2A participates in a fine control of pre-mature B-cell
apoptosis mediated by B-cell receptor signaling via
transcriptional regulation of survivin, IAP2 and
caspase-8 genes
Kenji Toyonaga
1,2,
*, Hidehiko Kikuchi
1,3,
*, Koki Yamashita
1
, Masami Nakayama
1
, Kazuo Chijiiwa
2
and Tatsuo Nakayama
1,3
1 Section of Biochemistry and Molecular Biology, Department of Medical Sciences, Miyazaki Medical College, University of Miyazaki, Japan
2 Section of Surgical Oncology and Regulation of Organ Function, Department of Medical Science, Miyazaki Medical College, University of
Miyazaki, Japan
3 Department of Life Science, Frontier Science Research Center, University of Miyazaki, Japan
It is widely known that B lymphocytes are essential to
immune responses in health and disease, and maintain
homeostasis by balancing cell viability and cell death
[1]. The molecular mechanisms of B-lymphocyte devel-
opment have mostly been studied in mammalian bone
marrow and/or peripheral lymphoid tissue (e.g. spleen).
Their development requires not only controlled lineage-
and locus-specific immunoglobulin gene recombination,
establishing unique antigen specificity of the B lympho-
cytes, but also developmental stage-specific gene expres-
sion participating in lymphoid cell proliferation and

tion, is used to study intracellular signal transduction of B lymphocytes.
Here, by analyzing the E2A-deficient DT40 cell line, E2A
)/)
, we show that
E2A deficiency prevents certain levels of apoptotic cell death mediated by
BCR signaling. In addition, E2A deficiency-linked BCR signaling controls
the mimicked pre-mature B-cell apoptosis by PMA/ionomycin through ele-
vated survivin plus inhibitor of apoptosis 2 levels, and reduced caspase-3
and caspase-8 activities, resulting in increased amounts of ICAD (inhibitor
of caspase-activated DNase), compared with those in the presence of E2A,
followed by reduction of DNA fragmentation. These findings will contrib-
ute to the resolution of molecular mechanisms of negative selection of
B cells and also auto-immune diseases.
Abbreviations
AIF, apoptosis-inducing factor; AKT, acutely transforming retrovirus AKT8 in rodent T cell lymphoma; BCR, B-cell receptor; CAD, caspase-
activated DNase; EBF, early B-cell factor; FACS, fluorescence-activated cell sorter; GAPDH, glyceraldehyde 3-phosphate dehydrogenase;
GATA-3, GATA binding protein-3; HAT, histone acetyltransferase; HDAC, histone deacetylase; IAP, inhibitor of apoptosis; PARP, poly(ADP-
ribose)polymerase; PCAF, p300/CBP-associated factor; PMA, phorbol 12-myristate 13-acetate.
1418 FEBS Journal 276 (2009) 1418–1428 ª 2009 The Authors Journal compilation ª 2009 FEBS
basic helix-loop-helix transcription factors E2A, early
B cell factor (EBF), GATA-binding protein-3 (GATA-
3), Pax5, PU.1, Ikaros and Aiolos, etc [2–4]. Their
importance in B-cell development has been established
by knockout experiments on mouse hematopoietic stem
cells, in which all these functions are associated with
cessation of early stages of B-cell differentiation [3,5].
However, the expression of these factors is sustained
throughout development of lymphocytes in normal
mice after the observed block in their knockout mice
[6]. Thus, their physiological functions beyond develop-

tor cells [19]. However, the impact of B cell-specific
factors, including E2A, on the BCR-mediated apop-
tosis of pre-mature B cells remains unclear, when
apoptosis is triggered by antigen stimulation.
In order to better understand the roles of histone
acetyltransferases (HATs), histone deacetylases
(HDACs) and B cell-specific transcription factors in
B-cell functions, we have systematically generated vari-
ous homozygous mutants, including HDAC2
)/)
,
GCN5
)/)
, Aiolos
)/)
and E2A
)/)
, using gene-targeting
techniques on the DT40 cell line, which was established
from chicken pre-mature B lymphocytes [20]. Our
results show that HDAC2 controls the amount of IgM
H-chain at two stages: transcription of its gene and
alternative processing of its pre-mRNA [21]. Recently,
we revealed not only that HDAC2 upregulates gene
expression of EBF1, Pax5, Aiolos, Ikaros and HDAC7,
and down-regulates those of E2A, p300/CBP-associated
factor (PCAF), HDAC4 and HDAC5, but also that
E2A upregulates expression of IgM H- and L-chain
genes, but downregulates Aiolos, but Aiolos, EBF1,
Pax5, and Ikaros downregulate expression of these two

expression of apoptosis-related factors
To assess the influence of E2A deficiency on gene
expression of apoptosis- and BCR signaling-related fac-
tors, and other factors, we performed semi-quantitative
RT-PCR on total RNAs prepared from DT40 and three
independent E2A
)/)
clones (Fig. 1). E2A deficiency did
not have a significant influence on transcription of most
of these genes, except for survivin (to approximately
200%), PKCa (to approximately 60%), PKCg (to
approximately 40%), PKCl (to approximately 40%)
and PKCf (to approximately 160%). In addition, we
performed immunoblot analyses to assess the influence
K. Toyonaga et al. Fine control of pre-mature B-cell apoptosis by E2A
FEBS Journal 276 (2009) 1418–1428 ª 2009 The Authors Journal compilation ª 2009 FEBS 1419
of E2A deficiency on the amounts of proteins whose
mRNA levels were altered as noted above. Consistent
with the results on mRNA levels, the protein levels for
survivin and PKCf were increased in E2A
)/)
and that
of PKCg was decreased (Fig. S1); PKCa and PKCl
could not be detected using the available antibodies. On
the other hand, transcription of various genes encoding
membrane-proximal factors, NF-jBs, transcription fac-
tors and B cell-related factors, amongst others, was not
altered in the E2A-deficient mutants (data not shown).
These insignificant effects of E2A deficiency on the
expression of numerous genes probably resulted in no

effects of PMA/ionomycin treatment on the viability
Fig. 1. Effect of E2A deficiency on gene expression of apoptosis-related factors, caspases, caspase-regulating factors, CAD/ICAD and PKCs.
Total RNAs were extracted from DT40 and three independent E2A
)/)
clones (1–3), and mRNA levels were determined by semi-quantitative
RT-PCR using appropriate primers. The chicken GAPDH gene was used as an internal control. The numbers under the panels indicate the
number of cycles used for PCR.
Fine control of pre-mature B-cell apoptosis by E2A K. Toyonaga et al.
1420 FEBS Journal 276 (2009) 1418–1428 ª 2009 The Authors Journal compilation ª 2009 FEBS
of DT40 and E2A
)/)
(Fig. 2B). As expected, the viabil-
ity of the two cell lines did not differ in the absence of
PMA/ionomycin. However, in the presence of PMA
and ionomycin, the viability of E2A
)/)
(approximately
50% at 48 h) was slightly higher than that of DT40,
which was dramatically reduced (approximately 20%
by 48 h). We undertook a comparative analysis of
changes in the morphological structure of nucleus as
an effect of PMA/ionomycin treatment in DT40 and
E2A
)/)
(Fig. 2C). Nuclear fragmentation, another
characteristic of apoptosis, was partially hindered in
E2A
)/)
, but it was clearly detected for DT40 in the
presence of PMA/ionomycin. These results show that

1
,G
1
, S and G
2
/M) for for DT40
and E2A
)/)
(clone 1), together with those of two other E2A
)/)
clones (2 and 3) are indicated in the table. (B) Sensitivity of DT40 (circles) and
E2A
)/)
(squares, triangles and diamonds) to PMA/ionomycin-mediated apoptotic cell death. Cells were resuspended in DMEM containing
10% v/v fetal bovine serum, and treated with (filled symbols) or without (open symbols) 10 ngÆmL
)1
PMA plus 1 lM ionomycin at 37 °C for
up to 48 h. Viable cells were counted by the trypan blue dye exclusion method. Data represent the mean of two separate experiments, and
error bars indicate the standard deviation. (C) Morphology of DT40 and E2A
)/)
(clone 1) cells treated with PMA/ionomycin. Cells were
cultured for 24 h without (no treatment) or with PMA/ionomycin (PMA/ionomycin), and their nucleus forms were analyzed by microscopy.
K. Toyonaga et al. Fine control of pre-mature B-cell apoptosis by E2A
FEBS Journal 276 (2009) 1418–1428 ª 2009 The Authors Journal compilation ª 2009 FEBS 1421
cooperatively induce apoptotic cell death in the DT40
cell line [24]. As a first step in elucidating the participa-
tion of E2A in apoptosis of DT40 cells, we examined
the effects of PMA/ionomycin treatment for these
homozygous DT40 mutants by FACS after staining
with propidium iodide and/or determination of cell

cin treatment dramatically altered gene expression of
E2A in DT40 (to approximately 320% by 24 h) and
that of Aiolos in both DT40 and E2A
)/)
(to less than
10% by 24 h). However, in the two cell lines, the treat-
ment did not have significant effects on mRNA levels
of HDAC1, HDAC2, SIRT2, HAT1 and MORF, and
slightly distinct but almost similar effects on mRNA
levels of EBF, Pax5, HDAC7, SIRT1, GCN5, PCAF
and MOZ. These findings, together with those shown
in Table 1, indicate not only that HDAC2 and GCN5
are necessary for control of the apoptosis of the DT40
cell line mediated by PMA/ionomycin treatment, but
also that their own transcription is not influenced by
PMA/ionomycin. Therefore, the apoptotic cell death
of DT40 mediated by PMA/ionomycin treatment must
be under the control of the elevated or decreased
amounts of E2A or Aiolos.
Upregulation of survivin and IAP2 gene expres-
sion and no effect on caspase-8 gene expression
by PMA/ionomycin treatment in E2A
-/-
To further clarify the molecular mechanism linked to
the apoptotic induction of the DT40 cell line coopera-
tively mediated by E2A and BCR signaling, DT40 and
E2A
)/)
were cultured in the presence of PMA/ionomy-
cin, and RT-PCR was performed for various factors

)/)
[22] ›
Aiolos
)/)
[22] fl
EBF
)/)
[22] fi
Helios
)/)
Unpublished data fi
Pax5
)/)
[22] fi
HDACs
HDAC1
)/)
[21] fi
HDAC2
)/)
[21] ›
HDAC7
)/)
[23] fi
SIRT1
)/)
[31] fi
SIRT2
)/)
[31] fi

effects on expression of caspase-3, caspase-6, caspase-
8, caspase-9 and caspase-10 genes (also shown in
Fig. 1). On the other hand, in both DT40 and E2A
)/)
,
the caspase-6 mRNA level was decreased gradually
by PMA/ionomycin treatment by 24 h, and the cas-
pase-10 mRNA level was increased by 3 h and there-
after decreased dramatically by 24 h. Expression of
caspase-3 and caspase-9 remained unchanged in the
presence of PMA/ionomycin. Interestingly, in DT40,
PMA/ionomycin treatment increased the caspase-8
mRNA level by 3 h (to approximately 160%) and
this level remained unchanged at 24 h, but the treat-
ment showed no change in the transcript level of
caspase-8 in E2A
)/)
. These findings indicate that
expression of most caspase genes is not much influ-
enced by either E2A or BCR stimulation, except that
of caspase-8.
With regard to caspase-regulating factors, depletion
of E2A increased transcription of the survivin gene (to
approximately 220%), but did not have a significant
effect on expression of the FLIP, IAP1, IAP2 and
Smac genes. However, PMA/ionomycin treatment had
distinct effects on expression of these caspase inhibi-
tors. The IAP1 or Smac mRNA levels increased or
decreased slightly by 3 h and thereafter remained
unchanged in both DT40 and E2A

downregulates transcription of the survivin gene and
has no effects on that of FLIP, IAP1, IAP2 or Smac,
but also that BCR stimulation and E2A cooperatively
control expression of FLIP, IAP2 and survivin genes.
Resistance to PMA/ionomycin-mediated
apoptosis of E2A
-/-
is brought about by
increased amounts of survivin and IAP2, and
reduced activity of caspase-3
Next we examined the effect of PMA/ionomycin treat-
ment on cellular protein levels of survivin, IAP2 and
ICAD, which are proximal factors controlling CAD
activity for DNA fragmentation, by immunoblotting
using their specific antibodies (Fig. 4A). E2A deficiency
increased the protein levels of survivin (to approxi-
mately 190%), but had no effect on those of IAP2 and
ICAD. Consistent with previous results [24], PMA/ion-
omycin treatment in DT40 dramatically decreased the
protein levels of IAP2 and ICAD (to approximately
40% and less than 10%) by 24 h, but had a moderate
influence on that of survivin (approximately 60% at
24 h). Therefore, in DT40 cells treated with PMA/iono-
mycin, the time courses of alterations in the protein and
mRNA levels of survivin were virtually similar, and the
gradual reductions in protein levels of IAP2 plus ICAD
compared with acute decreases (unchanged thereafter)
of their mRNA levels agreed with previous results [24].
On the other hand, in E2A
)/)

70% by 16 h), compared with those in DT40, probably
due to a balance of the amounts of each of the three
caspases and the inhibitors survivin and IAP2 (and also
FLIP and IAP1). The slightly decreased caspase-8
activity mediated by PMA/ionomycin treatment in
E2A
)/)
may have resulted from the balanced mRNA
(and probably protein) levels of caspase-8, which lead
to activation of pro-caspase-9, causing formation of the
active form of caspase-3, and of the inhibitors FLIP
(for caspase-8), survivin (for caspase-9) and IAP2 (for
caspase-3) [25]. In E2A
)/)
, the decreased activity of
caspase-3 mediated by PMA/ionomycin treatment must
depend on both the reduced amount of activated
caspase-3 itself as a result of decreased caspase-8 (and
probably caspase-9) activity, and the elevated protein
(and also mRNA) levels of the inhibitors survivin and
IAP2. As a result, the slow diminution of the protein
level of ICAD by PMA/ionomycin treatment in E2A
)/)
(Fig. 4A) must be due to its slight degradation medi-
ated by suppressed caspase-3 activity, although the
ICAD mRNA level was decreased when exposed for
3 h but thereafter remained unchanged, as did that in
DT40 (Fig. 3 and Fig. S2).
Because the alterations in the mRNA level of CAD
were the same in both DT40 and E2A

)/)
at indicated times
up to 24 h, and subjected to SDS–PAGE followed by immunoblotting. Antibody binding was detected using secondary antibodies conjugated
to horseradish peroxidase, and then data analysis was performed using a luminescent image analyzer. Left panel: typical immunoblot pattern
(DT40 and E2A
)/)
clone 1). b-actin was used as a control. The apparent molecular masses of marker proteins are indicated. Right panel: time
courses of protein levels for survivin, IAP2 and ICAD after treatment with PMA/ionomycin in DT40 (circles) and three E2A
)/)
clones (1–3)
(squares, triangles and diamonds). Data are expressed as percentages of the control (DT40 at 0 h). (B) Effects of PMA/ionomycin treatment
on caspase activities. Cell lysates were prepared from PMA/ionomycin-treated DT40 (circles) and E2A
)/)
clones (1–3) (squares, triangles and
diamonds) at indicated times up to 16 h, and then caspase activity assays were performed using appropriate caspase assay kits. Absorbance
at 405 nm was measured to determine activities. Data represent the mean of two separate experiments, and error bars indicate standard
deviation. (C) Effects of PMA/ionomycin treatment on DNA fragmentation in DT40 and E2A
)/)
. DNA was isolated from DT40 and E2A
)/)
cells incubated for 0, 8, 16 and 24 h in the presence of PMA and ionomycin, and analyzed by 1.5% agarose gel electrophoresis. The sizes
of k-DNA digested with HindIII are indicated in kb. Left panel: typical electrophoregram of DNA extracted from PMA/ionomycin-treated DT40
and E2A
)/)
(clone 1). Right panel: electrophoregram of DNA extracted from PMA/ionomycin-treated DT40 and three E2A
)/)
clones (1–3)
at 16 h.
K. Toyonaga et al. Fine control of pre-mature B-cell apoptosis by E2A
FEBS Journal 276 (2009) 1418–1428 ª 2009 The Authors Journal compilation ª 2009 FEBS 1425

Lack of E2A partially prevents the apoptotic cell
death seen in DT40 cells treated with PMA/ionomycin,
which mimics BCR stimulation (Fig. 2); such apoptosis
is completely prevented by either GCN5 deficiency [24]
or HDAC2 deficiency, and is significantly accelerated
by Aiolos deficiency (Table 1). By analyzing E2A
)/)
,we
revealed that E2A upregulates the expression of PKCa,
PKCg and PKCl genes, and downregulates the expres-
sion of survivin and PKCf genes, among the numerous
factors examined (Fig. 1 and unpublished data). PMA/
ionomycin treatment increased expression of the E2A
gene and dramatically suppressed that of the Aiolos
gene in DT40, but in E2A
)/)
had no effects or similar
effects on the expression of other disrupted genes
(Fig. 3A) and genes encoding B cell-specific factors,
HDACs and HATs (our unpublished data). These
results suggest not only that, among the B cell-specific
factors tested, E2A or Aiolos participates preferentially
in suppression or acceleration of apoptosis of the DT40
cell line, but also that, among the HAT and HDAC
families tested, gene expression of GCN5 and HDAC2,
which are essential for apoptotic cell death, are not
influenced directly by PMA/ionomycin treatment. On
the other hand, in E2A
)/)
, the detected protein level of

, the noticeable alterations in gene expression of
survivin and IAP2 accompanied by unchanged gene
expression of caspase-8 (and also various B cell-specific
factors, HDACs, HATs, apoptosis-related factors,
caspases, caspase-regulating factors and CAD/ICAD)
mediated by PMA/ionomycin results in suppression of
activities of caspase-3, caspase-8 and caspase-9, and
alterations in the protein levels of survivin, IAP2 and
ICAD (and probably FLIP) (Fig. 4A,B).
Finally, in E2A
)/)
, the slightly suppressed degrada-
tion of ICAD molecules as an effect of reduced cas-
pase-3 activity reduces CAD activity, leading to
moderate fragmentation of DNA molecules (Fig. 4C).
Thus, progress towards apoptotic cell death in E2A
)/)
is suppressed by collaboration of both BCR signaling
and E2A depletion, mainly via moderate changes in
amounts of the inhibitors survivin, IAP2 and ICAD
(and probably FLIP). Thus, E2A is involved in fine
control of pre-mature B-cell apoptosis mediated by
BCR signaling via transcriptional regulation of survi-
vin, IAP2, FLIP and caspase-8 genes.
The observations in this study regarding the par-
ticipation in apoptosis of Aiolos and HDAC2, which
are now being further studied by us, as well as that
of GCN5, amongst others, will be useful in elucidat-
ing not only the linkage between BCR signaling and
Fine control of pre-mature B-cell apoptosis by E2A K. Toyonaga et al.

)1
etoposide at 37 °C. Viable cells were counted
by the trypan blue dye exclusion method. Flow cytometric
analyses, morphological analyses, the caspase activity assay
and the DNA fragmentation assay were performed as
described previously [24,30].
Semi-quantitative RT-PCR
Total RNAs were isolated from DT40 and its subclones.
Reverse transcription was performed using a first-strand
DNA synthesis kit (Toyobo, Osaka, Japan) at 42 °C for
20 min, followed by heating at 99 °C for 5 min. PCRs were
performed as described previously [24] using sense primers
and antisense primers synthesized according to the EST
data deposited in GenBank for the appropriate genes, and
listed in previous reports [22–24], except for SIRT1 (sense
primer 5¢-CTGTTTTTACCACCAAATCG-3¢ and antisense
primer 5¢-CAACTTGTTGCTTGTTGGAT-3¢) and SIRT2
(sense primer 5¢-ATGTCCCTCATGGGCTTCGG-3¢ and
antisense primer 5¢-TCACGGCTCTTTGTCGTCCC-3¢).
The chicken glyceraldehyde 3-phosphate dehydrogenase
(GAPDH) gene was used as an internal control. PCR prod-
ucts were subjected to 1.5% agarose gel electrophoresis,
and analyzed using an LAS-1000plus luminescent image
analyzer (Fujifilm, Tokyo, Japan).
Immunoblotting
Cells were treated with 10% trichroloacetic acid, collected
by centrifugation 20 000 g for 5 min at 4 °C, dissolved in
0.5 m Tris/HCl (pH 6.8) containing 2.5% SDS, 10% glyc-
erol and 5% 2-mercaptoethanol, and heated at 100 °C for
5 min. Immunoblotting was performed as described previ-

Germany.
7 Riley RL, Blomberg BB & Frasca D (2005) B cells,
E2A, and aging. Immunol Rev 205, 30–47.
8 Ikawa T, Kawamoto H, Wright LYT & Murre C
(2004) Long-term cultured E2A-deficient hematopoietic
progenitor cells are pluripotent. Immunity 20, 349–360.
9 Lazorchak AS, Wojciechowski J, Dai M & Zhuang Y
(2006) E2A promotes the survival of precursor and
mature B lymphocytes. J Immunol 177, 2495–2504.
10 Lietz A, Janz M, Sigvardsson M, Jundt F, Dorken B &
Mathas S (2007) Loss of bHLH transcription factor
E2A activity in primary effusion lymphoma confers
resistance to apoptosis. Br J Haematol 137, 342–348.
K. Toyonaga et al. Fine control of pre-mature B-cell apoptosis by E2A
FEBS Journal 276 (2009) 1418–1428 ª 2009 The Authors Journal compilation ª 2009 FEBS 1427
11 Murre C (2005) Helix-loop-helix proteins and lympho-
cyte development. Nat Immunol 6, 1079–1086.
12 Rivera R & Murre C (2001) The regulation and func-
tion of the Id proteins in lymphocyte development.
Oncogene 20, 8308–8316.
13 Mikkers H, Allen J & Berns A (2002) Proviral activa-
tion of the tumor suppressor E2a contributes to T cell
lymphomagenesis in EmuMyc transgenic mice. Onco-
gene 21, 6559–6566.
14 Engel I & Murre C (1999) Ectopic expression of E47 or
E12 promotes the death of E2A-deficient lymphomas.
Proc Natl Acad Sci USA 96, 996–1001.
15 Carey GB, Donjerkovic D, Liu S, Hinshaw JA, Ton-
netti L, Davidson W & Scott DW (2000) B-cell receptor
and Fas-mediated signals for life and death. Immunol

23 Kikuchi H, Takami Y & Nakayama T (2005) GCN5: a
supervisor in all-inclusive control of vertebrate cell cycle
progression through transcription regulation of various
cell cycle-related genes. Gene 347, 83–97.
24 Kikuchi H & Nakayama T (2008) GCN5 and BCR sig-
nalling collaborate to induce pre-mature B cell apopto-
sis through depletion of ICAD and IAP2 and activation
of caspase activities. Gene 419, 48–55.
25 Aggarwal BB, Bhardwaj U & Takada Y (2004) Regula-
tion of TRAIL-induced apoptosis by ectopic expression
of antiapoptotic factors. Vitam Horm 67, 453–483.
26 Kee BL, Quong MW & Murre C (2000) E2A proteins:
essential regulators at multiple stages of B-cell develop-
ment. Immunol Rev 175, 138–149.
27 Lazorchak A, Jones ME & Zhuang Y (2005) New
insights into E-protein function in lymphocyte develop-
ment. Trends Immunol 26, 334–338.
28 Greenbaum S, Lazorchak AS & Zhuang Y (2004) Dif-
ferential functions for the transcription factor E2A in
positive and negative gene regulation in pre-B lympho-
cyte. J Biol Chem 279, 45028–45035.
29 Schwartz R, Engel I, Fallahi-Sichani M, Petrie HT &
Murre C (2006) Gene expression patterns define novel
roles for E47 in cell cycle progression, cytokine-medi-
ated signaling, and T lineage development. Proc Natl
Acad Sci USA 103, 9976–9981.
30 Kikuchi H & Imajoh-Ohmi S (1995) Activation and
possible involvement of calpain, a calcium-activated
cysteine protease, in down-regulation of apoptosis of
human monoblast U937 cells. Cell Death Differ 2,

Fine control of pre-mature B-cell apoptosis by E2A K. Toyonaga et al.
1428 FEBS Journal 276 (2009) 1418–1428 ª 2009 The Authors Journal compilation ª 2009 FEBS