The Transcription Factor GABP Is a Critical
Regulator of B Lymphocyte Development
Hai-Hui Xue,1,3Julie Bollenbacher-Reilley,1Zheng Wu,1Rosanne Spolski,1Xuefang Jing,3
Yi-Chen Zhang,1,4J. Philip McCoy,2and Warren J. Leonard1,*
1Laboratory of Molecular Immunology
2Flow Cytometry Core Facility
National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland 20892-1674
3Department of Microbiology, Carver College of Medicine, University of Iowa, Iowa City, Iowa 52242
4Present address: Harvard University, Cambridge, Massachusetts 02138.
GA binding protein (GABP) is a ubiquitously
expressed Ets-family transcription factor that
critically regulates the expression of the inter-
leukin-7 receptor a chain (IL-7Ra) in T cells,
whereas it is dispensable for IL-7Ra expression
in fetal liver B cells. Here we showed that defi-
ciency of GABPa, the DNA-binding subunit of
GABP, resulted in profoundly defective B cell
development and a compromised humoral im-
mune response, in addition to thymic develop-
mental defects. Furthermore, the expression
of Pax5 and Pax5 target genes such as Cd79a
was greatly diminished in GABPa-deficient B
cell progenitors, pro-B, and mature B cells.
GABP could bind to the regulatory regions of
Pax5 and Cd79a in vivo. Thus, GABP is a key
tional interaction of multiple transcription factors during
lineage specification and subsequent differentiation. Early
B cell development is critically dependent on at least five
transcription factors, including PU.1, Ikaros, E2A, early B
cell factor (EBF), and Pax5. Both PU.1 and Ikaros act in
parallel pathways at the level of lymphoid progenitors
(Bartholdy and Matthias, 2004; Busslinger, 2004; Hagman
and Lukin, 2005; O’Riordan and Grosschedl, 2000). The
generation of the earliest B cell progenitors requires E2A
and EBF, which coordinately activate B-lineage gene ex-
pression and initiate rearrangements at the immunoglob-
ulin heavy-chain gene locus at the onset of B-lymphopoi-
esis (Lin and Grosschedl, 1995; Zhuang et al., 1994). E2A
can directly activate transcription of the EBF gene (Kee
and Murre, 1998), and both E2A and EBF are required
for the activation of Pax5 gene expression (O’Riordan
and Grosschedl, 1999). Pax5 activates the expression of
B lineage signaling molecules and simultaneously re-
presses the transcription of lineage-inappropriate genes
(Nutt et al., 1997, 1998). These transcription factors thus
form an intricate gene-regulatory network that regulates
B-lineage specification and further commitment by acti-
vating and enforcing the expression of early B cell genes
(Medina et al., 2004).
The survival and proliferation of lymphoid progenitors in
fetal liver and bone marrow depend on cytokines, includ-
ing stem cell factor, Flt-3 ligand, and interleukin-7 (IL-7)
(Baird et al., 1999). IL-7 is primarily produced by stromal
cells in the thymus and bone marrow, and its receptor
consists oftheIL-7receptor achain(IL-7Ra)and thecom-
IL-7R system is absolutely required for the development
of T cells in humans and both B and T lymphocytes in
mice (Leonard, 2001). The expression of IL-7Ra in lym-
phoid progenitors and pro-B cells appears to be regulated
by PU.1 (DeKoter et al., 2002), whereas in mature T cells
where PU.1 is not expressed, IL-7Ra is transactivated
by another Ets factor, GABP (Xue et al., 2004). GABP is
ubiquitously expressed and critically regulates genes
that are necessary for cellular respiration in mitochondria,
cell-cycle control, and protein synthesis, as well as more
tissue-specific target genes (Rosmarin et al., 2004). Of
the more than 30 Ets family proteins, GABP is the only
one that functions as a heterodimer, consisting of a and
b chains that mediate DNA binding and transcriptional ac-
tivation, respectively (Sharrocks, 2001), whereas other Ets
developing mouse embryos, and inactivation of both
ski et al., 2004).
By using a gene-trap methodology, we have generated
mice in which a reporter gene was inserted in intron 5 of
the Gabpa allele, and in the embryos that are homozygous
for Gabpa-‘‘trapped’’ alleles (Gabpatp/tp), GABPa expres-
sion was greatly diminished but still detectable. With an
antiserum to the N terminus of GABPa, neither a shorter,
C-terminally truncated GABPa protein nor a longer fusion
protein of GABPa with the reporter gene was detected.
The hypomorphic expression of GABPa enabled the
Gabpatp/tpembryos to survive until E12.5 to E14.5, thus
Immunity 26, 421–431, April 2007 ª2007 Elsevier Inc. 421
allowing analysis of the role(s) of GABP in vivo (Xue et al.,
2004). By injecting Gabpatp/tpembryo-derived fetal liver
(FL) cells into Rag2?/?mice, we have now further investi-
with our previous observation that GABP critically regu-
lates IL-7Raexpression in T cells (Xue et al., 2004), the de-
velopment of thymocytes in Gabpatp/tpFL-reconstituted
Rag2?/?mice was severely compromised as anticipated.
Further, we found unexpected key roles of GABP in B cell
development, maturation, and function.
Defective T and B Cell Development in the Bone
Marrow of Gabpatp/tpFL-Reconstituted
To investigate the role of GABPa in lymphoid develop-
ment, we injected FL cells from E12.5 embryos into
Rag2?/?mice, which lack mature T and B cells. Lympho-
cytes that develop in the FL-reconstituted Rag2?/?mice
are therefore all donor derived. Consistent with the obser-
vations that Gabpa+/?mice had no detectable phenotype
and unaltered protein expression (Ristevski et al., 2004),
we found that both wild-type (WT, +/+) and heterozygous
FL cells had similar GABPa expression, and only
Gabpatp/tpFL cells showed diminished protein expression
(Figure S1 in the Supplemental Data available online).
When WT or Gabpa+/tpFL cells were transferred into
Rag2?/?mice, the cellularity in thymus and the number
of splenic T and B cells were increased (Figure 1A). Con-
sistent with the role of GABP in regulating IL-7Ra expres-
sion in T cells (Xue et al., 2004), the thymic cellularity was
greatly diminished in Gabpatp/tpFL-reconstituted Rag2?/?
mice (Figure 1A). In addition, there was a severe block
in thymocyte development in Gabpatp/tpFL-reconstituted
mice,witha marked increase
CD4?CD8?cells (Figure S2). Consistent with these find-
ings, splenic T cell numbers were also decreased (Fig-
ure 1A), but unexpectedly so were splenic B cell numbers
et al., 2004), in this manuscript, we focus on the unantici-
pated role of GABP in B cell development and function.
In the bone marrow, B220+IgM?cells include both pro-
B and pre-B lymphocytes, which develop into immature B
(Hardy and Hayakawa, 2001). In the bone marrow, B cell
development in Rag2?/?mice is blocked at pro-B cell
stage (Shinkai et al., 1992). In WT or Gabpa+/tpFL-recon-
stituted mice, surface IgM+recirculating and immature B
cells developed. In contrast, Gabpatp/tpFL-reconstituted
Rag2?/?mice had decreased numbers of pro-B and
pre-B cells in bone marrow, with an even more dramatic
decrease in the development of immature and recirculat-
ing B cells (Figure 1B).
cells, and B cell development in Rag2?/?mice is also
blocked at pro-B stage, the B220+IgM?cells in Gabpatp/tp
FL-reconstituted mice in Figure 1B presumably contained
both host- and donor-derived cells. To distinguish the or-
igin of these pro-B cells, we injected E13.5 FL cells that
express CD45.2 into Rag2?/?mice on the SJL.B6 back-
ground, which express the CD45.1 congenic marker. In
WT FL-reconstituted mice, B220+cells constituted about
15% of total bone-marrow cells, and almost all of these
B220+cells expressed CD45.2, indicating effective recon-
stitution with the donor hematopoietic stem cells (Fig-
ure 1C, top left). These B220+CD45.2+cells manifested
the expected progression from CD43+pro-B cells to
CD43?pre-B and immature B cells (Figure 1C, bottom
left). In contrast, Gabpatp/tpFL cells were much less effec-
tive in reconstituting the B cell compartment in the bone
marrow, with B220+CD45.2+cells accounting for only ap-
proximately 0.5% of total bone-marrow cells (Figure 1C,
top right). Like B220+CD45.2?cells in the Rag2?/?host,
most of the Gabpatp/tpdonor-derived B220+CD45.2+cells
expressed CD43, indicating an early block at the pro-B
cell stage (Figure 1C, bottom right). As quantified in
Figure 1D, far fewer early B precursors were generated
from Gabpatp/tpFL than from WT FL controls. We further
tested the ability of bone-marrow cells to generate pre-B
cells by using a pre-B cell colony-formation assay. Fewer
(only 0 to 3) colonies were generated from Gabpatp/tp
FL-reconstituted bone marrow than from WT FL (>20
colonies), even when 2- to 5-fold more cells were used
(Figure 1E), further corroborating that GABPa deficiency
caused a severe early block at the pro-B stage during B
Defective B Cell Maturation in the Spleens
of Gabpatp/tpFL-Reconstituted Mice
During B cell development, B cell precursors emigrate
from the bone marrow to undergo further maturation in
the spleen (Loder et al., 1999). In Gabpatp/tpFL-reconsti-
was observed. In mice with the more severe phenotypes
[denoted tp/tp (1)], almost no mature T and B cells ap-
peared in the spleen (Figure 2A), whereas in those with
less severe abnormalities [denoted tp/tp (2)], splenic B
cells were detected albeit at greatly lower percent and
numbers than in WT mice (Figures 1A and 2A). The range
of severity of defects in T and B cell maturation in different
mice might reflect different levels of leaky GABPa expres-
sion in donor FL cells. In our further analysis of B cell
maturation, we focused on the reconstituted mice with
mature B cells [tp/tp (2)-like] given the paucity of B cells
in tp/tp (1)-like mice. The Gabpatp/tpsplenic B cells ex-
pressed IgM and IgD, although the percent of IgM?IgD?
cells was higher than seen in the reconstitution experi-
ments with WT cells (Figure 2B). Further fractionation of
B220+cells (Hardy and Hayakawa, 2001) revealed that
marginal zone (MZ) B cells (CD21highCD23dim) appeared
at a roughly comparable frequency in the spleens of WT
and Gabpatp/tpFL-reconstituted mice (Figure 2C), al-
though the absolute number of Gabpatp/tpMZ B cells was
mature B cells in the B220+CD23?population (Xue et al.,
2003) were also found at a percentage comparable to
Role of GABP in B Cell Biology
422 Immunity 26, 421–431, April 2007 ª2007 Elsevier Inc.
controls, but again at a decreased absolute number (Fig-
ures 2D and 2E, middle). In contrast, B220+CD23+cells,
which include transitional 2 (T2) immature B cells and
follicular B cells (Hardy and Hayakawa, 2001), were
diminished in Gabpatp/tpFL-reconstituted mice (Figures
2C and 2E, right; two different Gabpatp/tpmice are
shown in Figure 2C to demonstrate the range of defects
in follicular B cell generation). Thus, GABPa is also re-
quired for B cell maturation in the periphery, especially
during the transition from T1 to T2 and then to follicular
Gabpatp/tpB Cells Have Diminished Function
To investigate whether the mature Gabpatp/tpB cells are
functional, we tested their proliferative response to mito-
genic stimulation by monitoring the dilution of carboxy-
fluorescein diacetate succinimidyl ester (CFSE). Although
Gabpatp/tpB cells responded to lipopolysaccharide (LPS),
anti-IgM, or anti-CD40 by upregulating CD40, B7.1, and
B7.2 expression (Figure 3A and data not shown), this
induction was moderately diminished, correlating with a
lower proliferative response of Gabpatp/tpB cells to stimu-
lation with LPS, as evaluated by a slower rate of CFSE
dilution than in B cells from WT FL-reconstituted mice
(Figure 3B). In Gabpatp/tpFL-reconstituted mice that
had mature T and B cells in the periphery, serum concen-
trations of IgM, IgG2a, IgG2b, and IgG3 were similar
to those in mice reconstituted with WT or Gabpa+/tpFL,
whereas IgG1 was markedly diminished (Figure 3C). This
is consistent with partial B cell function in Gabpatp/tp
mice as tested above in vitro, despite greatly diminished
B cell numbers.
Both B1 and marginal zone B cells are involved in the
early humoral response to T cell-independent antigens
(Martin et al., 2001), whereas follicular B cells participate
later in the T cell-dependent antibody response (Martin
and Kearney, 2002). To test the integrity of B cell function
in Gabpatp/tpFL-reconstituted mice in vivo, we immunized
nitropenol (TNP)-Ficoll, or a T cell-dependent antigen, ov-
strong antibody responses to both antigens, producing
antigen-specific IgM and IgGs (Figures 3D and 3E;
Figure S3). In contrast, Gabpatp/tpFL-reconstituted mice
responded to both antigens by generating antigen-spe-
cific IgM, albeit at lower amounts; however, the antigen-
specific IgG subtypes were markedly diminished (Figures
3D and 3E; Figure S3), indicating that GABPa is required
for normal immunoglobulin production. Given that T cell
development and function were also compromised in the
Figure 1. Impaired T and B Cell Develop-
(A) GABP is required for normal lymphoid
cellularity. Rag2?/?mice were reconstituted
with E12.5 FL cells from WT, Gabpa+/tp, or
Gabpatp/tpembryos. Total thymic cellularity
as well as splenic T and B cell numbers were
(B) Early block in B cell development in the
bone marrow of Gabpatp/tpFL-reconstituted
mice. Bone-marrow cells from Rag2?/?mice
or those reconstituted with WT or Gabpatp/tp
FL cells were stained with antibodies to B220
and IgM and analyzed by flow cytometry.
Gated populations are shown as the percent
of total bone-marrow cells, and immature B,
recirculating B, pro-B, and pre-B populations
are indicated. Shown are data representative
of five independent experiments.
(C) B cell development was blocked at pro-B
stage in Gabpatp/tpFL-reconstituted mice.
Rag2?/?mice on a B6.SJL background were
reconstituted with E13.5 FL cells, and the
bone-marrow cells were stained with anti-
bodies to B220, CD43, IgM, and CD45.2 and
analyzed as in (B). Pro-B and pre-B popula-
tions are indicated. Shown are data represen-
tative of three independent experiments.
(D) Comparison of the absolute numbers of
different B cell subpopulations in the bone
(E) Pre-B colony-formation assay. The colony
numbers are derived from 1 3 105bone-mar-
row cells. CFU, colony-forming unit. The data
are pooled results from three independent
Immunity 26, 421–431, April 2007 ª2007 Elsevier Inc. 423
Role of GABP in B Cell Biology
Gabpatp/tpFL-reconstituted mice (Figure 1A; Figure S2),
the defective humoral response to the T-dependent anti-
gen is likely due to the lack of proper T cell help in addition
to intrinsic B cell defects that are indicated by the defec-
tive response to the T-independent antigen.
Hematopoietic stem cells with lymphocyte differentiation
potential seed the incipient liver bud at fetal day E10,
as day E11 in fetal liver (de Andres et al., 2002). Although
the absolute FL cellularity was decreased in Gabpatp/tp
embryos, the B220+fraction that contains B cell precur-
sors was present at comparable frequencies in Gabpatp/tp
and control fetal livers (Figure 4A). To further study the
mechanism underlying the B cell developmental defects
resulting from GABPa deficiency, we sorted both line-
age-negative (Lin?) and B220+FL cells from E12.5 and
E13.5 embryos. As expected, in all populations, GABPa
expression was diminished (Figures 4B–4E), whereas the
expression of GABPb1 in B220+FL cells was not dimin-
ished and if anything slightly increased (Figures 4D and
4E). We first analyzed the expression of a set of
Figure 2. Impaired B Cell Maturation in Spleen of Gabpatp/tp
(A) Mature splenic T and B cells. Splenocytes were stained with anti-
bodies to CD3 and B220, and the percent of T and B cells is shown.
(B) IgM and IgD expression on B220+splenocytes.
(C) MZ and follicular B cells in the spleen. B220+splenocytes were
stained for CD21 and CD23. MZ (CD23dimCD21high) and follicular
(CD23bright) B cells are shown as the percent of B220+splenocytes.
(D) B220+splenocytes were stained for CD21, CD23, and IgM, and
B220+CD23dimcells were fractionated based on expression of CD21
and IgM. MZ (IgM+CD21high) and T1 (IgM+CD21medium) subpopulations
are shown as the percent of B220+CD23dimsplenocytes.
(E) Comparison of the absolute numbers of different splenic B cell
subpopulations. MZ, follicular, and T2 B cell numbers were based on
(C) and T1 B cell numbers were based on (D). The data in (A)–(D) are
representative data from at least five independent experiments.
Figure 3. Impaired Gabpatp/tpB Cell Functions
(A) Increased surface expression of B7.2 on LPS-stimulated splenic B
cells. B7.2 expression was determined on B220+splenocytes before
or 2 days after LPS stimulation, and the MFI is shown.
(B) Defective LPS-induced B cell proliferation in Gabpatp/tpFL-recon-
stituted Rag2?/?mice. Splenic B cells were stimulated with LPS, and
CFSE dilution was determined by flow cytometry at the indicated
(C) Defective basal serum levels of IgG1 in Gabpatp/tpFL-reconstituted
mice. Sera were collected from Rag2?/?mice reconstituted with WT,
Gabpa+/tp, or Gabpatp/tpFL, and the concentration of each Ig subtype
was determined by ELISA.
response in mice reconstituted with Gabpatp/tpFL. The reconstituted
mice were immunized with either TNP-Ficoll (D) or ovalbumin (E),
and sera were collected 1 week later, and then antigen-specific anti-
body titers were determined by ELISA. The data in (A), (B), (D), and
(E) are representative of three independent experiments.
424 Immunity 26, 421–431, April 2007 ª2007 Elsevier Inc.
Role of GABP in B Cell Biology
transcription factors that are required for B cell develop-
ment. In both E12.5 and E13.5 Lin?FL cells, Ikaros,
PU.1, and E2A mRNAs were relatively abundant, whereas
EBF and Pax5 mRNAs were expressed at much lower
amounts (Figures 4B and 4C). We next examined B220+
fetal liver cells that are enriched in B cell precursors (Fig-
ures 4D–4G). Interestingly, the increased expression of
Pax5 seen in WT B220+FL cells as compared to Lin?cells
was markedly decreased in Gabpatp/tpB220+FL cells from
both E12.5 and E13.5 embryos (18.0% ± 2.9% of WT in
E12.5, 26.3% ± 7.4% of WT in E13.5) (Figures 4D and
4E, compare filled and open circles for Pax5). Although
the upregulation of EBF expression in E12.5 Gabpatp/tp
B220+FL cells was relatively normal at day E12.5 (Fig-
ure 4D), it was approximately 69% less at day E13.5
(Figure 4E). A defect in the expression of both Pax5 and
EBF is consistent with a previous report that Pax5 is re-
quired at the onset of B lymphopoiesis in the fetal liver
and that the expression of EBF is profoundly decreased
in E18.5 Pax5?/?embryos (Nutt et al., 1997). Here, we
found that in E12.5 Gabpatp/tpFL cells, Pax5 expression
was markedly decreased but EBF expression was normal,
although EBF expression declined 1 day later. Although
EBF is known to regulate Pax5 expression, the fact that
Pax5 expression declined in E12.5 Gabpatp/tpFL prior to
the decrease in EBF expression excludes the possibility
that the initial decline in Pax5 expression is secondary to
decreased expression of EBF and suggests instead that
Pax5 may be directly regulated by GABP.
To further analyze the defects in B cell development re-
sulting from GABPa deficiency, we measured the expres-
cell surface proteins, and signaling molecules in B cells
(Bartholdy and Matthias, 2004; Busslinger, 2004; Hardy
and Hayakawa, 2001). Among these, the regulation of
Cd79a (which encodes Ig-a) has been extensively studied
and is known to be under the control of E2A, EBF, and
et al., 2004), and Pax5 recruiting Ets factors including
GABP to the Cd79a promoter (Fitzsimmons et al., 1996;
Maier et al., 2003). Approximately 90% and 86% de-
creases in expression of Cd79a in Gabpatp/tpFL cells
were observed at days E12.5 and E13.5, respectively (Fig-
resent a combined effect of diminished expression of
Pax5, EBF, and GABPa itself. Expression of other Pax5
target genes such as Cd19 and Blnk (which encode CD19
and B cell linker) (Nutt et al., 1998; Schebesta et al., 2002)
was also impaired (Figures 4F and 4G). The expression of
Vpreb1 and Cd79b (which encode V-pre-B and Ig-b),
which appears to be under the control of both EBF and
Pax5 (Akerblad et al., 1999; Nutt et al., 1997; O’Riordan
and Grosschedl, 1999), was also diminished (Figures 4F
and 4G). In contrast, the mRNAs of c-kit, Bruton’s tyrosine
kinase (Btk), and Spleen tyrosine kinase (Syk) were not af-
fected in Gabpatp/tpFL cells. Although Rag2 expression
was normal, Rag1 mRNA expression was somewhat di-
minished in Gabpatp/tpFL cells (Figures 4F and 4G).
Pax5 is known to maintain the identity of B cells by sup-
pressing lineage-inappropriate genes, and in Pax5?/?but
not WT bone-marrow pro-B cells, a number of genes that
are specific to other lineages, such as Csf2ra (which
encodes macrophage colony-stimulating factor receptor
a chain, MCSF-R) and Mpo (which encodes myeloperox-
idase, MPO) in myeloid cells, Prf1 (which encodes per-
forin) in NK cells, and Ptcra (which encodes pre-T cell an-
tigen receptor a, preTa) in T cells (Nutt et al., 1999), were
transcribed. In B220+E13.5 fetal liver cells, although per-
forin and preTa mRNAs were not detectable, we found
increased expression of MCSF-R (1.46- ± 0.4-fold, n = 4)
and MPO (2.97- ± 1.06-fold, n = 4) mRNAs, consistent
with diminished Pax5 expression resulting from de-
creased expression of GABPa.
Diminished Pax5 and Pax5 Target Gene Expression
in Gabpatp/tpFL-Derived Bone-Marrow Pro-B
and Splenic B Cells
Although B220+FL cells were present at a similar percent-
age in both Gabpatp/tpand control embryos, the cells
Figure 4. Defective Expression of B
Lineage-Specific Genes in Gabpatp/tpFL
(A) Frequency of B lineage precursor cells in
WT and Gabpatp/tpfetal liver. Fetal liver cells
from E12.5 embryos were stained for B220
and c-kit expression.
B220+ckit?subpopulations are shown as the
percent of total fetal liver cells.
(B–G) Altered gene expression in Gabpatp/tpFL
cells. Lin?or B220+cells were purified from
E12.5 or E13.5 fetal liver, and total RNA was
extracted, reverse-transcribed, and subjected
to quantitative PCR analysis. Copy numbers
of each mRNA relative to b-actin are shown.
Median value for each group was shown as
a horizontal line. N.D., not detectable.
Immunity 26, 421–431, April 2007 ª2007 Elsevier Inc. 425
Role of GABP in B Cell Biology
comprising the B220+populations could potentially be
different in WT embryos versus those with diminished
GABP expression. To further substantiate that the
gene-expression changes are intrinsic to B lineage cells,
we sorted CD45.2+B220+CD43+pro-B cells from the
bone marrow of reconstituted B6.SJL Rag2?/?mice.
Whereas PU.1 mRNA expression was similar to the WT
level in Gabpatp/tppro-B cells, the expression of Pax5,
EBF, and their target genes was substantially diminished
in Gabpatp/tppro-B cells (Figure 5A). We also analyzed
expression of these genes in mature splenic B cells sorted
from Rag2?/?mice that were reconstituted with either WT
or Gabpatp/tpFL and again observed a similar pattern,
although the defect in Gabpatp/tpmature B cells was less
severe (Figure 5B) than in the bone-marrow samples
(Figure 5A). It is evident that the severity of altered gene
expression decreased in association with progressive B
cell development. The most severe defects were found
in the fetal liver cells, the most primitive B cell precursors,
where Pax5 mRNA was decreased by approximately 82%
CD19, Ig-a, and VpreB mRNAs were markedly diminished
(Figures 4F and 4G). We also observed a substantial
defect in Gabpatp/tpFL-derived pro-B cells, in which Pax5
expression was only approximately 25%–30% of the
wild-type, with expression of Cd19 and Cd79a markedly
decreased, whereas the defect in mature splenic B cells
was rather modest. A possible explanation for these dif-
ferences is that B lineage precursor cells with greater
leakiness of expression of GABPa protein are most likely
to progress to later developmental stages and to appear
in the periphery as mature B cells.
To further substantiate a direct correlation between
GABPa and Pax5 expression, we also transiently trans-
fected 220-8 pro-B cells with small interference RNA
(siRNA) to silence the expression of GABPa as described
previously (Xue et al., 2004). The siRNA constructs effi-
ciently diminished expression of GABPa, whereas PU.1
expression was if anything slightly increased (Figure 5C).
The Pax5 mRNA expression was decreased by ap-
proximately 50% in cells with diminished GABPa mRNA
expression, whereas EBF mRNA expression was not sub-
ent decrease in EBF mRNA expression was only evident
in cells with prolonged low expression of GABPa, such
as in Gabpatp/tpB220+E13.5 FL cells and B220+CD43+
pro-B cells in Gabpatp/tpFL-reconstituted mice.
GABP Binds to the 50-UTR in the Pax5 Gene In Vitro
Because the expression of Pax5 and Pax5 target genes
was diminished in E12.5 Gabpatp/tpFL cells and Pax5 ex-
pression was diminished by transfection of the GABPa
siRNA constructs, we hypothesized that Pax5 might be a
direct target of GABP. To investigate this, we isolated a
putative promoter fragment spanning the murine B cell-
specific transcription initiation site in exon 1A of the Pax5
gene (extending from ?1627 to +294), cloned it upstream
of a luciferase reporter gene (Busslinger et al., 1996), and
increase in activity, and a further increase was seen with
further 50deletion to ?145 (Figure S4), an endpoint that
was still 50to the TATA box sequence (Figure S5). Align-
ment of the mouse and human PAX5 promoter and exon
1A regions revealed high conservation of sequences
including multiple Ets-binding motifs (i.e., GGAA or
TTCC) (Figure S5). It was previously shown that the Ets-
binding motif in the Il7r promoter region (herein denoted
EtsIL-7Ra/prom) is capable of binding both GABP and PU.1
(DeKoter et al., 2002; Xue et al., 2004). Consistent with
this, electrophoretic mobility shift assays (EMSAs) that
used an oligonucleotide probe spanning EtsIL-7Ra/prom
and nuclear extracts from ABL5-8 pro-B cells yielded
a fast-migrating minor complex (C2) containing PU.1 and
a slow-migrating major complex (C1) containing GABP
Figure 5. Diminished Gabpa Expression Resulted in Defec-
tive Expression of Pax5 and Pax5 Target Genes in Pro-B
and Splenic B Cells
(A) Altered gene expression in Gabpatp/tpFL-derived pro-B cells.
CD45.2+B220+CD43+pro-B cells were sorted from bone marrow
from WT or Gabpatp/tpFL-reconstituted B6.SJL Rag2?/?mice. Total
RNA was extracted and reversed-transcribed, and each mRNA was
analyzed by quantitative PCR. Data are means ± SD (n = 3).
(B) Altered gene expression in Gabpatp/tpFL-derived mature B cells.
Splenic B cells were isolated from WT or Gabpatp/tpFL-reconstituted
Rag2?/?mice by cell sorting, and mRNAs were analyzed as in (A).
Data are means ± SD (n = 4).
(C) Decreased Pax5 expression in cells expressing GABPa siRNA.
220-8 pro-B cells were transfected with either pBS/U6 vector or siRNA
constructs (siGABPa) together with pEYFP-N1. Total RNA was
extracted from sorted enhanced yellow fluorescent protein (EYFP)-
positive cells, and mRNAs were analyzed as in (A). Data are means ±
SD (n = 3).
426 Immunity 26, 421–431, April 2007 ª2007 Elsevier Inc.
Role of GABP in B Cell Biology
(Figure 6A; the identities of these bands were previously
identified in Xue et al., 2004). Among the potential Ets-
binding motifs in the Pax5 promoter and exon 1A regions
of which were evaluated by EMSA), only the Ets motif at
+210 to +213 in the 50-untranslated region (denoted
EtsPax5/50-UTR) formed a complex (C1) that corresponded
to GABP (Figure 6A). EtsPax5/50-UTRalso formed a C2 com-
plex corresponding to PU.1 (Figure 6A and data not
shown). Mutation of the consensus core sequence TTCC
to GTCG abrogated both C1 and C2 complexes (Fig-
ure 6B). Moreover, the GABP complex was efficiently
competed by WT but not mutant unlabeled oligonucleo-
tides spanning EtsIL-7Ra/prom(Figure 6C). We also per-
formed DNA-affinity purification by incubating the B cell
be highly enriched by the WT concatemers but not by the
mutated ones (Figure 6D). These results thus indicate that
like EtsIL-7Ra/prom, EtsPax5/50-UTRcan bind to both GABP
and PU.1 in vitro.
GABP Binds to the Pax5 50-UTR and Cd79a Promoter
To determine whether GABP or PU.1 binds EtsPax5/50-UTR
region in vivo, we performed chromatin immunoprecipita-
tion (ChIP) withantisera against GABPa, GABPb1,or PU.1
on murine thymocytes and purified splenic B cells (Xue
et al., 2004). As a control, we used a 790 bp sequence
that spans exon 8 of the Pax5 gene and lacks GGAA or
TTCC sequences. By normalizing with the exon 8 control
region, we found that the Pax5 50-UTR region was en-
enrichment was observed with DNA from the thymus
(Figure 7A), indicating that binding of GABP to the Pax5
50-UTR is specific to B lineage cells. In contrast, a PU.1
antibody didnotenrichthe Pax550-UTR regionfromeither
thymic or splenic B cell DNA preparations (Figure 7A),
even though this antibody works well for chromatin
immunoprecipitation with the Cd79a gene (see below).
These results suggest that GABP binds to EtsPax5/50-UTR
in vivo, whereas PU.1 does not, and this is consistent
with the finding that the expression of Pax5 in Gabpatp/tp
B220+FL cells was greatly diminished in spite of normal
expression of PU.1 (Figures 4D and 4E). To further sub-
stantiate the binding specificity of GABP to the Pax5
50UTR region, we designed probe-primer sets that detect
genomic sequences located approximately 1.2 kb up-
stream or 1.6 kb downstream of the 50-UTR region. As
shown in Figure 7B, the enrichment of Pax5 50-UTR by
anti-GABPa was apparent, whereas neither the upstream
nor the downstream region was enriched.
We next tested the importance of the EtsPax5/50-UTRsite
for Pax5 gene expression by using a luciferase reporter
assay. The activity of the ?145 to +294 reporter construct
from Figure S4 was substantially diminished when
EtsPax5/50-UTRwas mutated (Figure 7C); nevertheless,
some activity remained. Thus, although other elements in-
are additionally important for Pax5 reporter activity, the
ChIP data indicate that GABP binds to the Pax5 gene,
which correlates with the diminished Pax5 expression in
the Gabpatp/tpfetal liver cells, bone-marrow pro-B cells,
and splenic B cells (Figures 4 and 5).
Because Ig-a expression was also decreased in the
Gabpatp/tpB220+cells, we used the same splenic B cell
ChIP samples that we used for Pax5 (Figure 7A) to exam-
ine GABP and PU.1 binding to the Cd79a gene in vivo
(Figure 7D). Previous studies demonstrated that Pax5
can recruit three Ets factors including Ets-1, Fli-1, and
GABP to the Cd79a promoter, assembling ternary com-
plexes in vitro (Fitzsimmons et al., 1996; Maier et al.,
2003). Interestingly, both anti-GABPa and anti-GABPb1
Figure 6. GABP Bindsto thePax5 50-UTR
(A) A probe spanning EtsPax5/50-UTRforms DNA-
protein complexes that comigrate with GABP
and PU.1-derived complexes. Nuclear extract
from ABL5-8 pro-B cells was incubated with
probes spanning EtsIL-7Ra/promor EtsPax5/50-UTR.
binds both GABP and PU.1
inB cell nuclearextracts
EtsPax5/50-UTRforms three complexes as shown
(B) Mutation of EtsPax5/50-UTRabrogated the for-
mation of GABP and PU.1 complexes on
EMSA. In the mutant probe, the ‘‘TTCC’’ in EtsPax5/50-UTRwas changed to ‘‘GTCG’’ and used in EMSA as in (A).
(C) WT but not mutant unlabeled oligonucleotide spanning EtsIL-7Ra/promabolished the formation of GABP and PU.1 complexes with EtsPax5/50-UTR. In
EMSAs with B cell nuclear extract and the EtsPax5/50-UTRprobe, oligonucleotides spanning WT or mutant EtsIL-7Ra/promat the indicated molar excess
were included as competitors. Formation of protein-DNA complex was measured by EMSA.
EtsPax5/50-UTRor WT EtsIL-7Ra/promwere generated by self-primed PCR as described (Xue et al., 2004). The enriched proteins were analyzed by immuno-
blotting with GABPa or PU.1 antibody. Pull-down assays with concatemers containing EtsIL-7Ra/promor without concatemers were used as positive and
negative controls, respectively. Shown are data representative of at least three independent experiments for (A), (B), and (C) and of two independent
experiments for (D).
Immunity 26, 421–431, April 2007 ª2007 Elsevier Inc. 427
Role of GABP in B Cell Biology
fragments, and moreover, anti-PU.1 also enriched the
same region (Figure 7D), suggesting that both GABP and
PU.1 contribute to the transcriptional regulation of the
Cd79a gene and therefore that in addition to Cd79a being
a Pax5 target gene, it is also directly dependent on Ets-
family proteins, including GABP.
GABPa is ubiquitously expressed and has multifaceted
lar respiration (Rosmarin et al., 2004; Sharrocks, 2001). Its
critical function(s) in vivo are underscored by the preim-
plantation lethality resulting from complete inactivation
of both Gabpa alleles (Ristevski et al., 2004). Interestingly,
GABPa expression is enriched in embryonic stem cells
and neural stem cells (Fortunel et al., 2003), and both
GABPa and GABPb1 are enriched in hematopoietic
stem cells (HSCs) (Ivanova et al., 2002), which suggested
that GABP might be involved in maintaining self-renewal
and pluripotency of stem cells (Fortunel et al., 2003). In
this study, we investigated the roles of GABP in produc-
tion of B lineage cells from HSCs, B cell function, and
The hypomorphic residual expression of WT GABPa
protein in Gabpatp/tpembryos enabled longer embryo
survival, allowing us to evaluate GABPa function in the im-
opment is unexpectedly blocked at the pro-B stage be-
cause of GABPa deficiency. We correlated this early
arrest of B cell development with defective expression of
Pax5, EBF, and their target genes such as Ig-a in
GABPa-deficient B lineage cells. 50-UTRs not only can af-
fect mRNA stability and translation efficiency, but alsocan
contain binding motifs for regulatory proteins (Hughes,
2006). We show that GABPa and GABPb can directly
bind to an Ets motif in the Pax5 50-UTR in vitro and in
vivo, and this site is required for optimal Pax5 reporter ac-
tivity;however, theprecise contribution ofthissite tothe B
and it is also possible that other GABP-binding sites in the
moter region may cooperate with GABP in transactivating
tion of GABP with Pax5 has been observed on the Cd79a
promoter (Fitzsimmons et al., 1996; Maier et al., 2003).
The requirements for Pax5 in fetal and adult B lympho-
poiesis were reported to be different in studies with
Pax5?/?mice, and Pax5 is required at the onset of B lym-
phopoiesis in the fetal liver (Nutt et al., 1997). We found
that in E12.5 Gabpatp/tpfetal liver cells, Pax5 expression
was significantly decreased whereas EBF expression
was normal, suggesting that Pax5 up to this early stage
may be regulated by GABP but not by EBF. However,
EBF expression decreased in E13.5 Gabpatp/tpFL cells,
as did EBF target genes including Vpre-B, Ig-b, and Blk.
Given that Pax5 is essential for maintaining B lineage
Figure 7. GABP Binds to the Pax5 and
Cd79a Genes In Vivo
(A) In vivo binding of GABP to the Pax5 50-UTR.
prepared from primary thymocytes (Thy) or
splenic B cells and was immunoprecipitated
with rabbit IgG or purified antibody to GABPa,
GABPb1, or PU.1. Enrichment of DNA frag-
ments containing EtsPax5/50-UTRwas deter-
mined by normalizing to the signal obtained
via a region spanning exon 8 of the Pax5 gene.
(B) Specific association of GABPa with the
Pax5 50-UTR region. ChIP was performed as
in (A), and enrichment of chromatin fragments
corresponding to an upstream region, down-
stream region, or the Pax5 50-UTR region was
assessed by real-time PCR with probe-primer
sets, as indicated in the figure.
(C) The EtsPax5/50-UTRis required for optimal
Pax5 reporter activity. The indicated reporter
constructs were transfected into 220-8 pro-B
48 hr later. In Mut-Ets50-UTR, the Ets-binding
motif ‘‘TTCC’’ was changed to ‘‘GTCG.’’
(D) GABP binds to the Ig-a gene in vivo. ChIP,
detection, and normalization were performed
as in (A), and the enrichment of Cd79a pro-
moter region by each antibody was deter-
mined. Data in (A), (C), and (D) are means ±
SD (n = 3), and data in (B) are pooled results
(means ± SD) from two experiments, with
each experiment done in duplicate.
428 Immunity 26, 421–431, April 2007 ª2007 Elsevier Inc.
Role of GABP in B Cell Biology
commitment, the decline in Pax5 expression evident in
E12.5 Gabpatp/tpFL cells and the subsequent decrease
in EBF expression are detrimental to B cell development
in the fetal liver. Because EBF is required for the induction
of Pax5 (Medina et al., 2004; O’Riordan and Grosschedl,
1999), it is possible that the decline in EBF that we ob-
served by E13.5 helps to maintain a state in which levels
of Pax5 are decreased. Interestingly, analysis of gene ex-
pression in Pax5?/?bone marrow by semiquantative PCR
revealed that EBF mRNA was decreased by approxi-
mately 5-fold (Fuxa et al., 2004). In addition, analysis of
an EBF promoter region has identified a Pax5-binding
site (Roessler et al., 2006), suggesting a positive regula-
tory circuit between EBF and Pax5. Thus, although it is
possible that GABP can directly regulate EBF expression
and/or that GABP and EBF cooperatively regulate Pax5
transcription, the decreased EBF mRNA levels in Gab-
patp/tpFL and pro-B cells can at least partially be ascribed
to the loss of Pax5 expression.
Besides the requirement for key transcription factors
such as EBF and Pax5 for B cell development, Ig-a- and
Ig-b-mediated signaling is required for progression
through the early stages of this process. Mice harboring
mutations or deletions in the cytoplasmic domain of both
Ig-a and Ig-b exhibit a complete block at the pro-B cell
stage of B cell development (Kraus et al., 2001; Reichlin
et al., 2001). The transcription of Ig-a is under complex
regulation by multiple factors including EBF, Pax5, and
GABP itself (Sigvardsson et al., 2002). Thus, the early
block at the pro-B stage in Gabpatp/tpFL-reconstituted
mice might result from the diminished expression of Pax5
specific factors. In the periphery, small numbers of B cells
matured and appeared in the periphery, presumably be-
cause of the leaky expression of low levels of GABP.
ing from the loss of Pax5 function during late B cell lym-
phopoiesis in thatthere ispreferential lossof mature follic-
ular B cells and decreased expression of CD23 (Horcher
et al., 2001). Functionally, Gabpatp/tpand Pax5-deficient
B cells both showed decreased induction of B cell activa-
tion markers such as B7.2 in response to LPS stimulation
and greatly reduced production of IgG1, but comparable
production of IgM (Horcher et al., 2001). These observa-
tions further support the model that GABP helps to main-
tain B cell function, at least in part by supporting Pax5 ex-
pression. It is noteworthy, however, that the serum levels
of IgG2a, IgG2b, and IgG3 were reduced in Pax5 condi-
tional knockout mice but were relatively normal in
Gabpatp/tpFL-reconstituted mice. Although the basis for
we hypothesize that a complete GABP knockout in B cells
would result in a more profound defect.
Collectively, our data indicate that in addition to its
role(s) in maintaining basic cellular functions, GABP di-
rectly controls the expression of lineage-specific compo-
nents related to the immune system. GABP not only con-
trols T cell development at least in part via its regulation of
IL-7Ra expression, but it also controls B cell lymphopoie-
sis, with GABP deficiency resulting in decreased expres-
sion of Pax5, Ig-a, and other Pax5 target genes as early
as E12.5. Our findings thus reveal that GABP acts as
a key regulator that contributes to the differentiation of
hematopoietic stem cells into the B cell lineage.
Mice and Fetal Liver Cell Transplantation
Gabpa+/tpmice on a 129 x C57BL/6 (B6) mixed genetic background
mice, on B6 or B6.SJL backgrounds, were all obtained from Taconic.
All mice were handled in accord with NIH and AALAC guidelines and
under an animal study protocol approved by the NHLBI Animal Care
and Use Committee.
Single-cell suspensions were prepared from E12.5 or E13.5 fetal
livers and injected intravenously into Rag2?/?mice that received 600
rad whole-body irradiation 3–4 hr prior to injection. The injected
animals were analyzed 6–8 weeks after transplantation.
and stained with fluorochrome-conjugated antibodies, as described
(Xue et al., 2004). All fluorochrome-conjugated antibodies were from
Pre-B Colony-Formation Assay
Bone-marrow (BM) cells were recovered from reconstituted Rag2?/?
mice. 1 3 105BM cells were mixed with 3 ml of methylcellulose
medium for murine pre-B colony assays (StemCell Technologies)
and incubated at 37?C in a humidified environment for 1 week and
then the colonies were counted.
Splenic B cells were purified by positive selection with B220 MicroBe-
ads (MiltenyiBiotec), suspended in PBS containing 0.5% BSA, stained
with 5 mM CFSE at 37?C for 10 min, washed 3 times in complete
medium, and cultured in the presence of different stimuli. For B cell
stimulation, 5 mg/ml LPS (Sigma), 10 mg/ml anti-IgM m chain (Jackson
ImmunoResearch Laboratories), or 5 mg/ml anti-CD40 (BD PharMin-
gen) was used. The cells were collected on different days as indicated,
and the dilution of CFSE was monitored by flow cytometry.
To measure the immune response to a T cell-dependent antigen, mice
mixed with Imject Alum (Pierce) as an adjuvant. To measure the im-
mune response to a T cell-independent antigen, mice were injected
intraperitoneally with 100 mg of TNP(20)-AECM-Ficoll (Bioresearch
Technologies). Serum was collected 7 days after immunization and
analyzed for antigen-specific antibodies.
Serum Ig Titration
Ig levels in serial dilutions of serum were measured by enzyme-linked
immunosorbent assay (ELISA) with antibody pairs specific for different
min or 10 mg/ml of TNP(9)-BSA (Bioresearch Technologies). The TMB
bance at 450 nm was measured in a microplate reader (Perkin Elmer).
Transient Transfection of 220-8 Pro-B Cells
with siRNA Constructs
were described (Xue et al., 2004). A total of 20 mg of plasmid DNA
Immunity 26, 421–431, April 2007 ª2007 Elsevier Inc. 429
Role of GABP in B Cell Biology
(either the vector or the siRNA constructs), along with 1 mg or pEYFP-
N1, were transfected into 220-8 pro-B cells by electroporation with
a GenePulser Xcell electroporator (Bio-Rad) with the voltage and ex-
tension at 300 V and 950 mF, respectively. EYFP-positive cells were
isolated by cell sorting 60 hr after transfection, and gene expression
in these cells were assessed by real-time PCR.
Real-Time PCR Analysis
For evaluation of gene expression in B-lineage precursor cells in FL,
Lin?cells were isolated from FL single-cell suspension with mouse lin-
eage marker depletion kit (Miltenyi Biotec), and B220+FL cells or
B220+CD43+bone-marrow cells were isolated by cell sorting. Total
RNA was isolated with the RNeasy mini kit (Qiagen) and reverse tran-
scribed with Quantitech Reverse Transcription kit (Qiagen). Gene-spe-
cific probes and primer sets were from the Taqman Gene Expression
Assay system (Applied Biosystems) or synthesized from Operon Bio-
technologies. The reaction was performed on a 7900 HT sequence de-
tection system (Applied Biosystems). The expression of b-actin was
used to normalize the template input, and the copy number of b-actin
was arbitrarily set at 106. The relative abundance of each target gene
was calculated based on individual DCT, where DCTis the difference
of threshold cycle (CT) values between the target gene and b-actin in
the same sample. When the relative gene-expression levels in WT
and Gabpatp/tpFL-derived pro-B and mature B cells (Figure 5) were
compared directly, the mean value of individual DCTs in the WT group
was set at 1, and the expression level in Gabpatp/tpgroup was
calculated as 2?(tp/tp DCT-mean of WT DCT).
EMSA and Pull-Down Assays
The preparation of nuclear extract from ABL5-8 pro-B cells and EMSA
were performed as described (Xue et al., 2004). The WT and mutant
oligonucleotide probe sequences were 50-AGAAGCTCTTTAGTTCCT
AC-30, respectively. The sequences of probe spanning WT or mutant
EtsIL-7Ra/promwere as described (Xue et al., 2004). An oligonucleotide
containing 2 tandem copies of WT or mutant Pax5 probe sequence
and its complementary oligonucleotide were used in self-primed PCR,
and the product was used to pull down the interacting protein in the
nuclear extract from ABL5-8 pro-B cells after the procedures as de-
scribed (Xue et al., 2004). The interacting proteins were eluted with
0.6 M KCl and after dialysis were resolved on SDS-PAGE followed
by immunoblotting with anti-GABPa (H180, SantaCruz) or anti-PU.1
Luciferase Reporter Constructs and Luciferase Assay
The 50upstream regulatory region of Pax5 was amplified by PCR and
clonedintopGL4.10vector (Promega). Mutations inEts-bindingmotifs
were introduced by the QuickChange site-directed mutagenesis kit
(Stratagene). The reporter constructs (16 mg) and pGL4.74 expressing
Renilla luciferase driven by a TK promoter (200 ng) were cotransfected
Pulser II (Biorad). 48 hr later, luciferase activities were determined with
the Dual-Luciferase Reporter Assay system (Promega).
Positively selected splenic B cells or total thymocytes (50 million each)
were crosslinked with formaldehyde, and the fragmented chromatin
was prepared and immunoprecipitated with 5 mg of rabbit IgG or puri-
fied antibodies against GABPa and GABPb1 or anti-PU.1 as described
(Xue et al., 2004). Input and immunoprecipitated DNA samples were
analyzed by real-time PCR with the following probe and primers:
Cd79a promoter: probe, 50-GTAGGCGGAGTTTGGAGTGGAATCTG-30;
primers, 50-GGCACGGCTGAACAGGAA-30and 50-GTGGCCGTAGCC
TTGAGGTA-30; Pax5 50-UTR: probe (+192/+285), 50-ACGGTGCCTT
CGGACGCTTTTTTT-30; primers, 50-CGAGAAGCTCTTTAGTTCCTTA
ATCAT-30and 50-GATCACTGAGCTGAAACTAAACGTTT-30; Pax5 up-
stream region: probe (?1105/?1045), 50-TTGTGGCGAAATCTGCT
CAGTG-30; primers, 50-TCCACAGCCACGTTTCCA-30and 50-GCACC
GGACATCGCAAAT-30; Pax5 downstream region: probe (+1873/
50- GCCATCCTGTTTTGCTCCTTAG-30and 50-AAGTGGGTTTCCTGC
CATTG-30; Pax5 exon 8: probe, 50-TCGGCCTGGAGAATCAAAGCAA
CAA-30; primers, 50-GGGTTGTGCTCGCCAAGT-30and 50-GCAAAGA
The copy number of each target in immunoprecipitated DNA sam-
ples was obtained based on standard curves established from input
DNA. For each sample, the ratio of test regions in Pax5 or Cd79a pro-
of each test region by GABPa, GABPb1, or PU.1 antibody was ob-
tained relative to that obtained with control rabbit IgG.
Five figures are available at http://www.immunity.com/cgi/content/
We thank K. Cui for assistance with chromatin immunoprecipitation
and other help, L. Samsel and A. Williams for cell sorting, J. Flores
for assistancewithtailvein injection, and C. Robinson for maintenance
liver cell transplantation and M. Schlissel and H.C. Morse III for provid-
ing 220-8 and ABL5-8 pro-B cells, respectively. We thank B.J.
Fowlkes, R.N. Germain, K.-T. Jeang, J.-X. Lin, H.C. Morse III, R.H.
Schwartz, and K. Zhao for critical comments. This research was sup-
ported by the Intramural Research Program of the National Heart,
Lung, and Blood Institute, NIH. Y.-C.Z. was a pre-IRTA fellow at NIH
Received: October 19, 2006
Revised: January 9, 2007
Accepted: March 16, 2007
Published online: April 19, 2007
Akerblad, P., Rosberg, M., Leanderson, T., and Sigvardsson, M.
(1999). The B29 (immunoglobulin beta-chain) gene is a genetic target
for early B-cell factor. Mol. Cell. Biol. 19, 392–401.
Baird, A.M., Gerstein, R.M., and Berg, L.J. (1999). The role of cytokine
receptor signaling in lymphocyte development. Curr. Opin. Immunol.
Bartholdy, B., and Matthias, P. (2004). Transcriptional control of B cell
development and function. Gene 327, 1–23.
Busslinger, M. (2004). Transcriptional control of early B cell develop-
ment. Annu. Rev. Immunol. 22, 55–79.
Busslinger, M., Klix, N., Pfeffer, P., Graninger, P.G., and Kozmik, Z.
(1996). Deregulation of PAX-5 by translocation of the Emu enhancer
of the IgH locus adjacent to two alternative PAX-5 promoters in a dif-
fuse large-cell lymphoma. Proc. Natl. Acad. Sci. USA 93, 6129–6134.
de Andres, B., Gonzalo, P., Minguet, S., Martinez-Marin, J.A., Soro,
P.G., Marcos, M.A., and Gaspar, M.L. (2002). The first 3 days of
B-cell development in the mouse embryo. Blood 100, 4074–4081.
DeKoter, R.P., Lee, H.J., and Singh, H. (2002). PU.1 regulates expres-
Fitzsimmons, D., Hodsdon, W., Wheat, W., Maira, S.M., Wasylyk, B.,
and Hagman, J. (1996). Pax-5 (BSAP) recruits Ets proto-oncogene
family proteins to form functional ternary complexes on a B-cell-spe-
cific promoter. Genes Dev. 10, 2198–2211.
Fortunel, N.O., Otu, H.H., Ng, H.H., Chen, J., Mu, X., Chevassut, T., Li,
X., Joseph, M., Bailey, C., Hatzfeld, J.A., et al. (2003). Comment on
‘‘‘Stemness’: transcriptional profiling of embryonic and adult stem
cells’’ and ‘‘a stem cell molecular signature’’. Science 302, 393.
430 Immunity 26, 421–431, April 2007 ª2007 Elsevier Inc.
Role of GABP in B Cell Biology
Fuxa, M., Skok, J., Souabni, A., Salvagiotto, G., Roldan, E., and Download full-text
Busslinger, M. (2004). Pax5 induces V-to-DJ rearrangements and
locus contraction of the immunoglobulin heavy-chain gene. Genes
Dev. 18, 411–422.
Hagman, J., and Lukin, K. (2005). Early B-cell factor ‘pioneers’ the way
for B-cell development. Trends Immunol. 26, 455–461.
Hardy, R.R., and Hayakawa, K. (2001). B cell development pathways.
Annu. Rev. Immunol. 19, 595–621.
Horcher, M., Souabni, A., and Busslinger, M. (2001). Pax5/BSAP
maintains the identity of B cells in late B lymphopoiesis. Immunity
Hughes, T.A. (2006). Regulation of gene expression by alternative
untranslated regions. Trends Genet. 22, 119–122.
Ivanova, N.B., Dimos, J.T., Schaniel, C., Hackney, J.A., Moore, K.A.,
and Lemischka, I.R. (2002). A stem cell molecular signature. Science
Kee, B.L., and Murre, C. (1998). Induction of early B cell factor (EBF)
and multiple Blineagegenesbythebasic helix-loop-helixtranscription
factor E12. J. Exp. Med. 188, 699–713.
Kraus, M., Pao, L.I., Reichlin, A., Hu, Y., Canono, B., Cambier, J.C.,
Nussenzweig, M.C., and Rajewsky, K. (2001). Interference with immu-
noglobulin (Ig)alpha immunoreceptor tyrosine-based activation motif
(ITAM) phosphorylation modulates or blocks B cell development, de-
pending on the availability of an Igbeta cytoplasmic tail. J. Exp. Med.
Leonard, W.J. (2001). Cytokines and immunodeficiency diseases.
Nat. Rev. Immunol. 1, 200–208.
Lin, H., and Grosschedl, R. (1995). Failure of B-cell differentiation in
mice lacking the transcription factor EBF. Nature 376, 263–267.
Loder, F., Mutschler, B., Ray, R.J., Paige, C.J., Sideras, P., Torres, R.,
Lamers,M.C., and Carsetti,R.(1999). Bcell development inthe spleen
takes place in discrete steps and is determined by the quality of B cell
receptor-derived signals. J. Exp. Med. 190, 75–89.
Maier, H., Ostraat, R., Parenti, S., Fitzsimmons, D., Abraham, L.J.,
Garvie, C.W., and Hagman, J. (2003). Requirements for selective re-
cruitment of Ets proteins and activation of mb-1/Ig-alpha gene tran-
scription by Pax-5 (BSAP). Nucleic Acids Res. 31, 5483–5489.
Maier, H., Ostraat, R., Gao, H., Fields, S., Shinton, S.A., Medina, K.L.,
Ikawa, T., Murre, C., Singh, H., Hardy, R.R., and Hagman, J. (2004).
Early B cell factor cooperates with Runx1 and mediates epigenetic
changes associated with mb-1 transcription. Nat. Immunol. 5, 1069–
Martin, F., and Kearney, J.F. (2002). Marginal-zone B cells. Nat. Rev.
Immunol. 2, 323–335.
Martin, F.,Oliver,A.M.,andKearney,J.F.(2001). MarginalzoneandB1
B cells unite in the early response against T-independent blood-borne
particulate antigens. Immunity 14, 617–629.
Medina,K.L., Pongubala, J.M., Reddy, K.L., Lancki, D.W., Dekoter, R.,
Kieslinger, M., Grosschedl, R., and Singh, H. (2004). Assembling
a gene regulatory network for specification of the B cell fate. Dev.
Cell 7, 607–617.
Nutt, S.L., Urbanek, P., Rolink, A., and Busslinger, M. (1997). Essential
functions of Pax5 (BSAP) in pro-B cell development: difference be-
tween fetal and adult B lymphopoiesis and reduced V-to-DJ recombi-
nation at the IgH locus. Genes Dev. 11, 476–491.
Nutt, S.L., Morrison, A.M., Dorfler, P., Rolink, A., and Busslinger, M.
Nutt, S.L., Heavey, B., Rolink, A.G., and Busslinger, M. (1999). Com-
mitment to the B-lymphoid lineage depends on the transcription factor
Pax5. Nature 401, 556–562.
O’Riordan, M., and Grosschedl, R. (1999). Coordinate regulation of B
cell differentiation by the transcription factors EBF and E2A. Immunity
O’Riordan, M., and Grosschedl, R. (2000). Transcriptional regulation
of early B-lymphocyte differentiation. Immunol. Rev. 175, 94–103.
Reichlin, A., Hu, Y., Meffre, E., Nagaoka, H., Gong, S., Kraus, M.,
Rajewsky, K.,and Nussenzweig,M.C. (2001).Bcelldevelopment isar-
rested at the immature B cell stage in mice carrying a mutation in the
cytoplasmic domain of immunoglobulin beta. J.Exp. Med. 193,13–23.
Ristevski, S., O’Leary, D.A., Thornell, A.P., Owen, M.J., Kola, I., and
tial for early embryogenesis. Mol. Cell. Biol. 24, 5844–5849.
Roessler, S., Gyory, I., Imhof, S., Spivakov, M., Williams, R.R.,
Busslinger, M., Fisher, A.G., and Grosschedl, R. (2006). Distinct pro-
moters mediate the regulation of Ebf1 gene expression by IL-7 and
Pax5. Mol. Cell. Biol. 27, 579–594.
Rosmarin, A.G., Resendes, K.K., Yang, Z., McMillan, J.N., and
Fleming, S.L. (2004). GA-binding protein transcription factor: a review
of GABP as an integrator of intracellular signaling and protein-protein
interactions. Blood Cells Mol. Dis. 32, 143–154.
Schebesta, M., Pfeffer, P.L., and Busslinger, M. (2002). Control of
pre-BCR signaling by Pax5-dependent activation of the BLNK gene.
Immunity 17, 473–485.
Sharrocks, A.D. (2001). The ETS-domain transcription factor family.
Nat. Rev. Mol. Cell Biol. 2, 827–837.
Shinkai, Y., Rathbun, G., Lam, K.P., Oltz, E.M., Stewart, V.,
Mendelsohn, M., Charron, J., Datta, M., Young, F., Stall, A.M., et al.
ity to initiate V(D)J rearrangement. Cell 68, 855–867.
Sigvardsson, M., Clark, D.R., Fitzsimmons, D., Doyle, M.,Akerblad,P.,
Breslin, T., Bilke, S., Li, R., Yeamans, C., Zhang, G., and Hagman, J.
(2002). Early B-cell factor, E2A, and Pax-5 cooperate to activate the
early B cell-specific mb-1 promoter. Mol. Cell. Biol. 22, 8539–8551.
Xue, L., Morris, S.W., Orihuela, C., Tuomanen, E., Cui, X., Wen, R., and
Wang, D. (2003). Defective development and function of Bcl10-defi-
cient follicular, marginal zone and B1 B cells. Nat. Immunol. 4, 857–
Xue, H.H., Bollenbacher, J., Rovella, V., Tripuraneni, R., Du, Y.B., Liu,
C.Y., Williams, A., McCoy, J.P., and Leonard, W.J. (2004). GA binding
protein regulates interleukin 7 receptor alpha-chain gene expression in
T cells. Nat. Immunol. 5, 1036–1044.
Zhuang, Y., Soriano, P., and Weintraub, H. (1994). The helix-loop-helix
gene E2A is required for B cell formation. Cell 79, 875–884.
Immunity 26, 421–431, April 2007 ª2007 Elsevier Inc. 431
Role of GABP in B Cell Biology