(Received for publication, June 11, 1996, and in revised form, November 14, 1996)
From the Shionogi Institute for Medical Science, 2-5-1 Mishima, Settsu-shi, Osaka 566, Japan
TAL1, which is frequently activated in T cell acute lymphoblastic leukemia (T-ALL), encodes lineage-specific basic helix-loop-helix (bHLH) proteins that bind specifically to E-box DNA motif upon dimerization with ubiquitous basic helix-loop-helix proteins E47 or E12. RBTN1 and RBTN2, also frequently activated in T-ALL, encode proteins only with tandem cysteine-rich LIM domains. We found that aberrant expression of TAL1 detected in 11 out of 14 T-ALL cell lines was invariably accompanied by that of either RBTN1 or RBTN2. Forced expression of TAL1 together with RBTN1 or RBTN2, but not TAL1 alone, strongly induced artificial reporter genes in a TAL1/RBTN-negative T-ALL cell line, HPB-ALL. Such collaborative transcriptional activity of TAL1 and RBTN was not, however, observed in non-T cell lines, suggesting further involvement of some T cell-specific cofactors. In this context, we carried out preliminary evaluation of a potential role of the T cell-specific GATA-binding protein, GATA3, in the transcriptional activity of TAL1 and RBTN. We also showed that coexpression of TAL1 and RBTN1 in HPB-ALL strongly induced TALLA1, a highly specific T-ALL marker whose positivity correlated 100% with ectopic expression of TAL1 among various T-ALL cell lines. Collectively, ectopic TAL1 and RBTN1 or -2, together with some endogenous T cell-specific cofactors like GATA3, constitute a highly collaborative set of transcription factors whose aberrant activity in T cells may lead to leukemogenesis by modulating expression of downstream genes such as TALLA1.
TAL1, also called SCL or TCL5, is
a gene whose aberrant activation in the T cell lineage by recurrent
chromosomal translocations, t(1;14)(p32;q11) and t(1;7)(p32;q35),
precise ~90-kilobase pair interstitial chromosomal deletions
(tald), and yet other undefined mechanisms is
implicated as the major pathway for the development of T cell acute
lymphoblastic leukemia (T-ALL)1 (1-3).
TAL1 is normally expressed in erythroid, mastocytic, and
megakaryocytic lineages of the hematopoietic system but not in T cells
(4, 5). TAL1 encodes at least two polypeptides, full-length
42-kDa TAL1 (amino acid residues 1-331) and 22-kDa N-terminally
truncated polypeptide TAL1
(amino acid residues 176-331) (6), both
containing the basic helix-loop-helix (bHLH) motif, a DNA-binding and
protein dimerization domain found in a number of transcription factors.
TAL1 proteins, having no intrinsic DNA binding activity, dimerize with
the ubiquitously expressed E2A gene products, E47 and E12
(7, 8), and the heterodimers bind to E-box elements (CANNTG) with a
preferred sequence of AA
GT (9). By using an
artificial reporter gene containing multiple copies of the optimal
TAL1/E2A binding sequence, transcriptional activity of the TAL1/E47
heterodimer was examined in transiently transfected murine C3H/10T1/2
fibroblasts (10). While the E47 homodimer strongly induced the reporter
gene, the TAL1/E47 heterodimer was much less active, suggesting a
negative regulatory role of TAL1 (10). So far, nothing is known about
genes regulated by TAL1. Transcriptional activity of TAL1 in T cells
has not been examined either. Transgenic mice expressing 42-kDa TAL1 in
the T cell lineage did not develop tumors nor did mice irradiated and
reconstituted with bone marrow cells infected with a recombinant retrovirus encoding TAL1 (11, 12). Such findings may suggest that
additional cofactors are necessary for TAL1 to be oncogenic in T cells
(13).
RBTN1/TTG1 and RBTN2/TTG2 are genes that also were originally identified from recurrent chromosomal translocations in T-ALL, t(11;14)(p15;q11) and t(11;14)(p13;q11), respectively (14-16). They encode highly related proteins consisting of only two cysteine-rich zinc finger-like LIM domains, which are considered to mediate protein-protein interactions (17). RBTN1 and RBTN2 are mainly expressed in the brain and erythroid lineage cells, respectively, but not in normal T cells (18, 19). Transgenic mice expressing RBTN1 or RBTN2 in T cells developed tumors but at variable frequencies and only after long latency periods (~10 months) (20, 21), again suggesting a need for cofactors.
In normal hematopoiesis TAL1 is considered to function in close collaboration with RBTN2 and a group of GATA binding zinc finger type transcription factors, GATA1 and GATA2 (19). This conclusion comes from their overlapping patterns of expression in developing erythroid cells (4, 5, 18, 22, 23), their proven physical interactions in vitro as well as in vivo (24-26), and a very similar abnormality in the embryonic erythropoiesis that was produced by disruption of each of these genes (27-30). It is thus conceivable that the activity of TAL1 in T cells as a transcriptional and oncogenic protein requires a similar set of transcriptional cofactors. Recently, efficient collaboration of TAL1 and RBTN2 in T cell oncogenesis has indeed been demonstrated in vivo by generating double transgenic mice (31).
Previously we identified a highly specific surface marker of T-ALL designated TALLA1 (32). TALLA1 is a new member of the tetraspans or transmembrane 4 superfamily and normally expressed in neurons, certain vascular endothelial cells, and certain epithelial cells but not at all in any hematopoietic cells including T cells (32).2 The highly frequent ectopic expression of TALLA1 in T-ALL suggests that its aberrant expression is closely related to leukemogenesis of T-ALL. We have not, however, detected any T-ALL-associated genetic changes involving the TALLA1 gene that is located on X chromosome.3 One possibility is that aberrantly activated transcription factors in T-ALL such as TAL1 are responsible for ectopic expression of TALLA1. In the present study, we examined expression of TAL1, RBTN1, and RBTN2 in a panel of 14 human T-ALL cell lines and evaluated transcriptional activity of TAL1 and RBTN by monitoring expression of transfected artificial reporter genes and the endogenous TALLA1 gene. The transcriptional activity of TAL1 indeed requires RBTN1 or RBTN2 together with some other T cell-specific cofactors and induces the TALLA1 gene. Such a highly collaborative set of transcriptional factors may thus lead to development of T-ALL by modulating expression of downstream genes like TALLA1.
All of the cell lines used in the present study were cultured in RPMI 1640 supplemented with 10% fetal bovine serum. For details of each cell line see Takagi et al. (32).
RT-PCRTotal RNA samples were prepared from various cell
lines using Trizol® reagent (Life Technologies, Inc.). The first
strand cDNA synthesis was carried out from 1 µg of total RNA in
20 µl of reaction buffer using a RNA PCR kit (Takara Shuzo, Kyoto,
Japan). PCR was then set up using 2 µl from the reaction buffer as
template in a 50-µl PCR reaction buffer containing 2.5 units of
Pfu DNA polymerase (Stratagene). Amplification was carried
out by denaturation at 94 °C for 1 min (5 min at the first cycle)
and annealing/extension at 70 °C for 5 min. The number of
amplification cycles was 30 for TAL1, RBTN1, and E2A; 27 for GATA3; and
35 for RBTN2. Amplified products were electrophoresed on a 1.5%
agarose gel and stained with ethidium bromide. The oligonucleotides
used as PCR primers were: TAL1,
+5-TCACCACCAACAATCGAGTGAAGAGG-3
and
5
-CTCCTCCTGGTCATTGAGCAGCTTGG-3
; RBTN1,
+5
-CGGAGCGCCCGAGATGATGGTGCTGG-3
and
5
-GCAGTCGAGGTGATACACGTTGTCCC-3
; RBTN2,
+5
-TCCCTCCCCAATGTCCTCGGCCATCG-3
and
5
-CTCGGGCCTATATCATCCCATTGATC-3
; E2A,
+5
-CAGCAGGGTTTCCAGGCCTGAGGTGC-3
and
5
-GCTGCTGTGCGACTCAGTGAAGTGGG-3
; GATA3,
+5
-TAAGATCGACGGTCAAGGCAACCACG-3
and
5
-GTGGTGGATGGACGCTTGGAGAAGG-3
.
The cDNAs for TAL1, RBTN1,
RBTN2, and GATA3 were amplified from Jurkat or CEM by RT-PCR. The
oligonucleotide primers used were as follows: TAL1,
+5
-GAGAGTCTAGACTCTCTAAATATGCCCCAGGATGACC-3
and
5
-GAGAGTCTAGAGCTGGATGCCTCAGATGAGAGCTGAC-3
; TAL1
,
+5
-GAGAGTCTAGAAGCCGGATGCCTTCCCTATGTTCACC-3
and
5
-GAGAGTCTAGAGCTGGATGCCTCAGATGAGAGCTGAC-3
; RBTN1,
+5
-CGGAGCGCCCGAGATGATGGTGCTGG-3
and
5
-GCGTTACTGAACTTGGGATTCAAAGG-3
; RBTN2,
+5
-TCCCTCCCCAATGTCCTCGGCCATCG-3
and
5
-CTCGGGCCTATATCATCCCATTGATC-3
; GATA3,
+5
-AGCACAGCCGAGGCCATGGAGGTGAC-3
and
5
-GTGAGCATCGAGCAGGGCTCTAACCC-3
. Amplified cDNAs were
cloned into pCRTMII (Invitrogen) and digested with EcoRI.
The EcoRI fragments were then cloned into pMIKneo and
pMIKhyg, yielding pMIKneoTAL1
, pMIKneoTAL1
, pMIKneoGATA3,
pMIKhygRBTN1, and pMIKhygRBTN2. pMIKneo and pMIKhyg, which carry the
genes conferring resistance to neomycin and hygromycin B, respectively,
were constructed from an efficient eukaryotic expression vector pSR
(33) and kindly provided by Dr. K. Maruyama at The Institute of Medical
Science, Tokyo University. To make a reporter plasmid tk-CAT, the
herpes simplex virus thymidine kinase (tk) minimal promoter
(34) was amplified by PCR using tk-F
(+5
-GACTGCAGTCAACACGCAGATGCAGTC-3
) and tk-R
(
5
-GACTGCAGGGTCGCTCGGTGTTCGAGG-3
), digested with PstI,
and cloned into the PstI site of pCAT®-Basic (Promega). To
make a reporter plasmid tk-TAL1CS-CAT, the double-stranded DNA
containing four copies of the optimal TAL1/E2A-binding motif (AACAGATGGT) (9) was prepared by annealing two synthetic
oligonucleotides, +5
-GATCCAGAAGCTT
CACACG
AAGCTTCTGGGG-3
and
5
-GATCCCCCAGAAGCTT
TCGTGTG
AAGCTTCTG-3
and cloned in tandem into the HindIII site of tk-CAT. The
dominant negative GATA3 KRR mutant (35) was kindly provided by Astar Winoto. Recipient cells (1 × 107) were transfected by
electroporation with indicated combinations of plasmids together with 5 µg of pRC/cytomegalovirus-luciferase carrying the luciferase gene
under the control of cytomegalovirus promoter. After 72 h cell
lysates were prepared and assayed for CAT activity after normalization
of transfection efficiency by luciferase activity as described
previously (36).
The EcoRI fragments of the RBTN1 cDNA and RBTN2 cDNA were subcloned into the EcoRI site of the maltose-binding protein (MBP) fusion vector pMAL-c (New England Biolabs). The EcoRI fragment of the GATA3 cDNA was subcloned into the EcoRI site of the glutathione S-transferase (GST) fusion vector pGEX-2T (Pharmacia Biotech Inc.). Production of fusion proteins (MBP-RBTN1, MBP-RBTN2, and GST-GATA3) in Escherichia coli BL21 (Novagen) and their purification were carried out following the protocols recommended by the manufacturers. GST alone or GST-GATA3 fusion protein was immobilized onto glutathione-Sepharose 4B (Pharmacia), washed with phosphate-buffered saline, and incubated at 4 °C for 1 h with MBP alone or MBP-RBTN fusion proteins that were preincubated with 1% skim milk in phosphate-buffered saline at 4 °C for 20 min. Beads were spun down to collect supernatants and washed with cold phosphate-buffered saline 5 times. Bound proteins were eluted with 5 mM reduced glutathione. The eluted proteins and supernatants were run on SDS-polyacrylamide gel electrophoresis and analyzed by immunoblotting using anti-MBP (New England Biolabs).
Isolation of Stable TransformantsHPB-ALL cells (1 × 107) were cotransfected by electroporation with pMIKneo or
pMIKneoTAL1 and pMIKhyg or pMIKhygRBTN1. Cells were seeded into
96-well plates at 5 × 105/ml and cultured under
double selection with 600 µg/ml G418 and 400 µg/ml hygromycin B. Rare growing cultures were individually expanded and determined for
expression of transfected genes by RT-PCR (see above). Surface
expression of TALLA1 was analyzed by flow cytometry on a FACStar Plus
(Becton Dickinson) after indirect immunofluorescence staining with
monoclonal anti-TALLA1 (B2D) as described previously (32). Northern
blot analysis for TALLA1 and glyceraldehyde-3-phosphate dehydrogenase
transcripts was also carried out as described previously (32).
Ectopic expression of the TAL1 gene in T-ALL is
now known to occur much more frequently than previously considered from
the incidence of gross chromosomal rearrangements involving the
TAL1 gene locus (2). Similarly, ectopic expression of
RBTN1 and RBTN2 in T-ALL may be more frequent
than the incidence of respective chromosomal translocations (3).
Therefore, we examined 14 T-ALL cell lines for expression of
TAL1, RBTN1, and RBTN2 together with that of E2A, whose products E47 and E12 are necessary for
the TAL1 gene products to form a DNA binding heterodimer
(7-10). As shown in Fig. 1, 11 out of 14 T-ALL cell
lines were found to express TAL1, supporting a highly
frequent aberrant expression of the TAL1 gene in T-ALL (2).
Importantly, the same 11 cell lines were found to be exactly the ones
that expressed either RBTN1 (2/11) or RBTN2
(9/11). The remaining 3 T-ALL cell lines expressed neither of them. All
of the 14 cell lines expressed the E2A gene, which was thus
useful as an internal control. The complete concordance of aberrant
expression of TAL1 and RBTN1 or RBTN2
in 11 out of 14 T-ALL cell lines was striking, supporting their
interdependent transcriptional and oncogenic activity in T cells. It is
also notable that the 11 T-ALL cell lines that coexpressed
TAL1 and RBTN1 or RBTN2 are exactly
the ones that are strongly positive for surface expression of TALLA1, a
highly specific tumor marker of T-ALL (32).
Transcriptional Activity of TAL1 Requires RBTN1 or RBTN2
No
natural target genes regulated by TAL1 are known yet. Hsu et
al. (9) have shown that the heterodimers of TAL1 and E47 or E12
specifically bind to an E-box motif in vitro and regulate expression of an artificial reporter gene containing multiple copies of
the optimal TAL1/E2A binding motif in murine C3H10T1/2 fibroblasts
(10). We therefore decided to use a similar artificial reporter gene
system to examine transcriptional activity of TAL1 and RBTN in T cells.
The HSV tk minimal promoter (34) was linked to the CAT
structural gene in pCAT-Basic (Promega), making tk-CAT, and then a DNA
segment containing four copies of the optimal TAL1/E2A-binding site
(AACAGATGGT) (9) was inserted into the upstream sequence of the
tk promoter, making tk-TAL1CS-CAT. HPB-ALL cells, which expressed E2A but not TAL1, RBTN1, or
RBTN2 (Fig. 1), were used as a recipient. Due to alternative
translation initiation sites, TAL1 proteins have at least two isoforms
containing the bHLH motif, the 42-kDa TAL1 and an N-terminally
truncated 22-kDa TAL1
(6). We therefore tested the activity of both
polypeptides.
As shown in Fig. 2A, TAL1 or TAL1
alone
did not significantly induce expression of the CAT reporter gene from
tk-TAL1CS-CAT. RBTN1 or RBTN2 had no effect on the expression of the
reporter gene either. In the presence of RBTN1 or RBTN2, however, TAL1 proteins strongly induced the reporter gene. The full-length TAL1
and the N-terminally truncated TAL1
were similarly effective in the
presence of RBTN1 or RBTN2. Thus, RBTN1, which is expressed in the
brain (18) and RBTN2, which is expressed in erythroid cells like TAL1
(19), both efficiently cooperated with TAL1. Unexpectedly, however, the
reporter plasmid with only the tk minimal promoter and
without 4 copies of the TAL1/E2A-binding motif (tk-CAT) was also nearly
half as much induced by the coexpression of TAL1 and RBTN as
tk-TAL1CS-CAT. Computer analysis revealed a single potential E-box
element in the HSV tk minimal promoter and as many as 12 potential E-box elements in the pCAT-Basic plasmid itself. Some of
these sequences may thus fortuitously function as a TAL1/E2A-binding
site. The deletion of the potential E-box element in the tk
promoter, however, did not affect induction of tk-CAT by the
combination of TAL1 and RBTN (data not shown). The role of any of the
E-box-like elements within pCAT-Basic in the induction by TAL1 and RBTN
remains to be seen.
We carried out the same transfection assays using a B-ALL cell line, BALL-1. As shown in Fig. 2B, little transcriptional activity of TAL1 proteins on tk-TAL1CS-CAT or tk-CAT was observed in BALL-1 even in the presence of RBTN1 or RBTN2. Similar negative results were obtained with another B cell line Raji or a monocytoid cell line U937 (data not shown). It is therefore likely that the transcriptional activity of TAL1 proteins requires, besides RBTN, still other cofactor(s) specifically present in T cells.
Effect of GATA3 on Transcriptional Activity of TAL1In the
erythroid lineage cells, GATA1 and GATA2, the zinc finger GATA binding
transcription factors are expressed at high levels (19), and in close
collaboration with TAL1 and RBTN2 they play essential roles in the
erythropoietic development and differentiation (27-30). Furthermore,
Osada et al. (24) demonstrated direct binding of RBTN2 with
TAL1 and GATA2. On the other hand, they observed little significant
binding between TAL1 and GATA1 or GATA2 (24). These results suggest
that RBTN2 mediates interactions between TAL1 and GATA proteins. Among
the GATA proteins, GATA3 is known to be expressed very specifically in
the T cell lineage (37). We confirmed this in Fig. 3
where RT-PCR detected the expression of GATA3 in all the
T-ALL cell lines tested but not in any other hematopoietic cell lines
including BALL-1. Since it was not known whether RBTN1 or RBTN2 was
capable of binding to the T cell-specific GATA3, we also examined
direct binding of RBTN1 and RBTN2 to GATA3. To do this, GATA3 was
expressed as a GST fusion protein, while RBTN1 and RBTN2 were expressed
as MBP fusion proteins. GST alone or GST-GATA3 was immobilized by glutathione beads and incubated with MBP alone, MBP-RBTN1, or MBP-RBTN2. Free proteins and bound proteins eluted by glutathione were
analyzed by immunoblot with anti-MBP. As shown in Fig.
4, GST-GATA3 but not GST alone retained MBP-RBTN1 and
MBP-RBTN2 but not MBP alone. Thus, both RBTN1 and RBTN2 are capable of
binding to GATA3. We also examined direct binding between TAL1 and
GATA3 by the same protocol using MBP-TAL1 fusion protein and GST-GATA3 fusion protein. We did not observe any direct binding between TAL1 and
GATA3 (data not shown).
These results led us to examine the effect of GATA3 on the
transcriptional activity of TAL1/RBTN in T cells and non-T cells. For
the sake of convenience, we used only the reporter plasmid tk-TAL1CS-CAT in the following experiments. We first examined the
effect of GATA3 overexpression on the transcriptional activity of
TAL1 in HPB-ALL (Fig. 5A). Transfection of
GATA3 had no direct effect on the expression of tk-TAL1CS-CAT. GATA3
also did not collaborate with TAL1
. On the other hand, GATA3 clearly
augmented the collaborative transcriptional activity of TAL1
and
RBTN1 or RBTN2 by 2-3-fold. We next carried out the same transfection experiment using BALL-1 (Fig. 5B). As already shown in Fig.
2B, coexpression of TAL1
and RBTN1 or -2 hardly induced
expression of tk-TAL1CS-CAT. The combination of TAL1
and GATA3 was,
however, found to induce the reporter gene to some extent even without RBTN1 or -2. The combination of TAL1, GATA3, and especially RBTN1 further enhanced expression of the reporter gene to some extent. It
should, however, be noted that the levels of induction of the reporter
gene in BALL-1 even by the full combination of TAL1, GATA3, and RBTN1
or -2 were still far less than those seen in HPB-ALL by the combination
of TAL1 and RBTN1 or -2. Therefore, the exact role of GATA3 as a
cofactor in the collaborative transcriptional activity of TAL1 and RBTN
needs further evaluation.
Induction of TALLA1 in HPB-ALL Stably Transformed with TAL1
As mentioned above, the aberrant expression of
TAL1 in T-ALL cell lines correlates 100% with strong
expression of TALLA1, a highly specific tumor marker of T-ALL not
detected on normal T cells (32). To examine the role of TAL1 and RBTN
in the ectopic expression of TALLA1, we cotransfected
HPB-ALL with the following combinations of expression plasmids: pMIKneo + pMIKhyg, pMIKneoTAL1 + pMIKhyg, pMIKneo + pMIKhygRBTN1, and
pMIKneoTAL1
+ pMIKhygRBTN1. Stable transformants were isolated by
double selection with G418 and hygromycin B, and expression of TAL1
and RBTN1 in each clone was determined by RT-PCR (data not shown). Two
representative clones from each combination were then examined for
surface TALLA1 by indirect immunofluorescence staining and flow
cytometry (Fig. 6A). HPB-ALL clones
transformed with vectors only or those expressing either TAL1
or
RBTN1 alone showed little surface expression of TALLA1. In sharp
contrast, HPB-ALL clones coexpressing TAL1
and RBTN1 were strongly
positive for TALLA1. The two clones coexpressing TAL1
and RBTN1 also
contained a large quantity of TALLA1 mRNA, whereas clones
transfected with vectors only or those expressing either TAL1
or
RBTN1 alone showed no such accumulation of TALLA1 mRNA (Fig.
6B). These results strongly suggest that the collaborative transcriptional activity of TAL1 and RBTN indeed induces, either directly or indirectly, the TALLA1 gene in T-ALL.
TAL1, TAL2, and LYL1, whose bHLH domains are most closely related to each other among >60 bHLH proteins so far described, are all identified as the genes activated in T-ALL by recurrent chromosomal translocations (38, 39). It is thus conceivable that these three bHLH proteins have a common mode of action in the leukemogenesis of T cells. RBTN1 and RBTN2, encoding closely related proteins containing only tandem LIM domains, represent another group of genes frequently activated by T-ALL-specific chromosomal translocations (14-16). Wadman et al. (26) reported that TAL1, TAL2, and LYL1 all interacted with RBTN1 and RBTN2. Furthermore, Larson et al. (31) recently reported that double transgenic mice expressing both TAL1 and RBTN2 in T cells developed T cell tumors much more efficiently than those expressing either gene alone (12, 20, 21). RBTNs, which are devoid of intrinsic DNA binding activity, may thus function as an essential cofactor for TAL1 and related bHLH proteins in their transcriptional and oncogenic activity in T cells. In the present study, we have demonstrated that the transcriptional activity of TAL1 in T cells indeed requires RBTN1 or RBTN2.
First, we found that all the T-ALL cell lines expressing TAL1 do coexpress either RBTN1 or RBTN2 with 100% concordance (Fig. 1). Second, forced expression of TAL1 in a T-ALL cell line HPB-ALL strongly activated transcription from an artificial reporter gene under the control of the HSV tk minimal promoter with four copies of the optimal TAL1/E2A binding E-box motif in the upstream sequence (9) only when RBTN1 or RBTN2 was coexpressed (Fig. 2). Third, TALLA1, a highly specific tumor marker of T-ALL (32) whose positivity shows 100% concordance with that of TAL1 among various T-ALL cell lines, was strongly induced by coexpression of TAL1 and RBTN1 in a TALLA1-negative T-ALL cell line HPB-ALL (Fig. 6). Collectively, RBTN1 or RBTN2 is indeed an essential cofactor for the positive transcriptional activity of TAL1 not only on the artificial reporter gene but also, either directly or indirectly, on an endogenous gene TALLA1 in T cells. Thus, the highly frequent ectopic expression of TALLA1 in T-ALL (32) strongly suggests the presence of aberrant collaborative transcriptional activity of TAL1 and RBTN in most T-ALL cases. The finding that mice coexpressing TAL1 and RBTN2 in the T cell lineage efficiently developed T cell tumors (31) further underscores the importance of such collaborative transcriptional activity of TAL1 and RBTN in triggering the oncogenic process of T cells. Studies are now in progress to examine how the TALLA1 gene is induced in T cells by TAL1 and RBTN.
Unexpectedly, the reporter gene with the HSV tk minimal promoter but without the upstream optimal TAL1/E2A-binding elements was also quite efficiently induced by the coexpression of TAL1 and RBTN1 or -2 in HPB-ALL (Fig. 2). We have not yet examined the mechanism of this induction in detail. However, we may point out that there are a number of potential E-box-like elements in the original CAT plasmid (pCAT-Basic), some of which may function as fortuitous TAL1-responsive elements. Alternatively, TAL1 and RBTN may affect, somehow in a highly T cell-specific manner (see below), the expression and/or activity of transcription factors such as Sp1 and CTF that are known to interact with the HSV tk minimal promoter (40). These possibilities are now under investigation.
Surprisingly, TAL1 even in the presence of RBTN1 or RBTN2 hardly induced the artificial reporter genes in non-T cells (Fig. 2). This suggests that the collaborative transcriptional activity of TAL1 and RBTN1 or -2 further requires some other cofactors specifically present in T cells. The fact that tumor-specific translocations involving TAL1, RBTN1, and RBTN2 are highly restricted to T-ALL and not observed in any other types of tumors (3) also supports that some T-cell specific cofactors are essential for the collaborative oncogenic activity of TAL1 and RBTN. In erythroid cells, TAL1 and RBTN2 are known to function in close collaboration with GATA1 or GATA2, the members of GATA-binding zinc finger type transcription factors (19). It is thus possible that the T cell-specific member of the GATA proteins, GATA3 (37) (Fig. 3), may be one of such cofactors in T cells. We showed that RBTN1 and RBTN2 were capable of directly binding to GATA3 in vitro (Fig. 4). We further demonstrated that overexpression of GATA3 in HPB-ALL augmented the collaborative transcriptional activity of TAL1 and RBTN1 or RBTN2 (Fig. 5). In BALL-1, TAL1 with and even without RBTN collaborated with GATA3 to induce the reporter gene (Fig. 5). These results may support that GATA3 is a cofactor for the collaborative transcriptional activity of TAL1 and RBTN1 or -2 in T cells. However, it should also be noted that the observed collaborative transcriptional activities of TAL1, GATA3, and RBTN1 or -2 in BALL-1 were far less than those seen by the combination of TAL1 and RBTN1 or -2 in HPB-ALL. Furthermore, we found that a dominant negative mutant of GATA3 KRR (35) had no inhibitory effect at all on the collaborative transcriptional activity of TAL1 and RBTN in HPB-ALL (data not shown). The KRR mutant, which has three amino acids KRR (305-307) in the region between the two zinc fingers mutated to alanines, was shown to be totally inactive on a reporter gene linked to multiple GATA3-binding sites and to block transactivation of the same reporter gene by wild type GATA1, GATA2, or GATA3 (35). However, the results with the KRR mutant still do not formally exclude the possibility of GATA3 as a cofactor because the KRR mutation may not affect the ability of GATA3 to function as such. Nevertheless, all of these observations may argue against a critical role of GATA3 in the collaborative transcriptional activity of TAL1 and RBTN1 or -2 in T cells. Thus, there may be more relevant cofactors for the TAL1/RBTN transcriptional complex in T cells besides or even independent of GATA3.
Obviously further studies are needed to define cofactors for TAL1 and RBTN in T cells and to identify downstream genes whose expression in T cells is modulated, like TALLA1, by the TAL1/RBTN transcriptional complex. Such studies will lead to better understanding of the leukemogenesis of T cells and to a new therapeutic strategy for T-ALL.
We thank Dr. K. Maruyama for pMIKneo and pMIKhyg, Dr. Astar Winoto for pc-KRR, Dr. S. Takagi for fluorescence-activated cell sorter analysis, and Dr. Yorio Hinuma and Dr. Masakazu Hatanaka for constant support and encouragement.