Transcriptional activation of the cardiac homeobox gene
CSX1/NKX2-5 in a B-cell chronic lymphoproliferative
Xinying Su,1,5Véronique Della-Valle,1,5Eric Delabesse,3Zahia Azgui,2Roland Berger,1
Hélène Merle-Béral,2,4Olivier A Bernard,1,5and Florence Nguyen-Khac1,2,4
1INSERM, E0210, Hôpital Necker, Paris;2Service d’Hematologie Biologique, Groupe Hospitalier Pitie-Salpetriere,
Paris;3Laboratoire d’Hématologie, CHU Purpan, Toulouse;4Université Pierre et Marie Curie, Paris VI;5Université
Rene Descartes, Paris V, France
Homeobox containing transcription factors are frequently deregulated in human hematologic malignant diseases either indirectly through
an abnormality of an upstream factor, or directly through rearrangement of the gene itself. Study of T-cell acute lymphoblastic leukemia
identified the related non-clustered homeobox transcription factors, TLX1 and TLX3, as frequently ectopically expressed as a result of
chromosomal translocations.We report the deregulation of a non-clustered homeobox gene in a new type of t(5;14)(q35;q11) translo-
cation in a mature peripheral B-cell leukemia.This translocation results in the ectopic expression of the CSX1/NKX2-5 gene on chromo-
some 5q35 due to its juxtaposition to the TCR δ gene on chromosome 14q11. Expression of the CSX1/NKX2-5 protein conferred
enhanced replating potential to transduced murine bone marrow cells. Our study establishes that deregulation of homeobox encoding
genes is not restricted to acute leukemic proliferations, but is also observed in chronic malignant diseases.
Key words: chromosomal translocation, homeobox, transcriptional activation, chronic lymphocytic leukemia.
Citation: Su CX,Della-Valle V,Delabesse E,Azgui Z,Berger R,Merle-Béral H,Bernard OA and Nguyen-Khac F.Transcriptional activation of
the cardiac homeobox gene CSX1/NKX2-5 in a B-cell chronic lymphoproliferative disorder. Haematologica 2008; 93:1081-1085.
©2008 Ferrata Storti Foundation.This is an open access paper.
Acknowledgments: we would like to thank H. Drabkin for critical reading of the manuscript, Dr. M. Schoenwald for the initial karyotype, and J. Ong and L.
Merlin for expert technical help.
Funding: this work was supported partially by INSERM and by the Comité de Paris de la Ligue Nationale Contre le Cancer (LNCC).
Manuscript received December 6, 2007. Revised version arrived on February 4, 2008. Manuscript accepted February 22, 2008.
Correspondence: Florence Nguyen-Khac, Service d’Hematologie Biologique Groupe Hospitalier Pitié-Salpetriére, Pavillon Laveran, 47-83 Bd de l'Hopital,
75013 Paris, France. E-mail: firstname.lastname@example.org.The online version of this article contains a supplementary appendix.
Chromosomal abnormalities observed in human malignant
diseases are a frequent mechanism of gene deregulation or
disruption. Chromosomal translocations involving the T-cell
receptor (TCR) and immunoglobulin (BCR) genes are exclu-
sively seen in lymphoid tumors.1Generally TCR gene translo-
cations are closely associated with T-cell lymphoid tumors,
whereas BCR translocations are associated with B-cell malig-
nant diseases. The studies of chromosomal translocations
allowed the identification of several human oncogenes that
can also be targeted by other types of oncogenic events, like
mutations. An example is the NOTCH1 gene, isolated in
humans because of its rearrangement with the TCR β chain
gene in the rare t(7;9)(q34;q34) translocation and now known
to be altered in more than 50% of T-cell acute lymphoblastic
One of the major targets in leukemogenesis is the family of
homeobox-containing genes which encode DNA-binding
proteins. They can be deregulated by various mechanisms,
directly or indirectly. Transcriptional deregulation of clustered
homeobox genes is observed as an indirect result of chromo-
somal translocations involving the MLL gene, a regulator of
the Trithorax family, or from yet uncharacterized causes in
various leukemic subtypes.3PBX1, an homeobox-cofactor of
HOX gene, is fused to the E2A gene by the t(1;19)(q23;p13)
translocation of B-cell ALL.4
The HOX11/TLX1 orphan gene family is inappropriately
expressed in about 30% of T-ALL samples.5-7TLX1, located on
chromosome 10q24, is ectopically expressed upon its
rearrangement with the TCR δ, or TCR β, chain genes in 10%
of T-ALLs. The related HOX11L2/TLX3 gene, located on chro-
mosome 5q35, is rearranged with the TCR δ gene on chromo-
some 14 (band q11) in only a few cases of ALL.8However, it
is targeted by the cryptic t(5;14)(q35;q32) translocation, which
affects the CTIP2/BCL11B locus on chromosome 14 (band
haematologica | 2008; 93(7) | 1081 |
X. Su et al.
| 1082 | haematologica | 2008; 93(7)
q32) in more than 20% of childhood T-ALLs.9,10Finally,
a new chromosome 5q35 breakpoint has been reported
close to the CSX1/NKX2-5 homeobox gene, approxi-
mately 2 megabases telomeric to TLX3, which juxta-
poses NKX2-5 to BCL11B locus derived sequences in
two cell lines11and to TCRD in a T-ALL patient.12
In a search for additional structural abnormalities of
the long arm of chromosome 5, we identified a
t(5;14)(q35;q11) translocation in a B-cell chronic lym-
phoproliferative disorder. We now report the molecular
characterization of this translocation and demonstrate
the involvement of the TCR δ gene on chromosome 14
and the NKX2-5 gene on chromosome 5.
A 49-year-old female was diagnosed with an atypical
B-lymphoproliferative disorder with peripheral blood,
bone marrow and breast involvement in the Pitié-
Salpêtrière hospital in October 1995. Spleen and lymph
nodes were not involved. The immunophenotype of
peripheral blood lymphocytes was CD22+, FMC7+,
CD79b+, CD23–, CD5+and IgM/D κ high (Matutes'
score 2).13Apart from CD5, no T-cell markers were
expressed. According to current WHO criteria,14this
case could be classified pleomorphic mantle cell lym-
phoma. After CHOP/CLL type treatment,15complete
remission was observed until May 2000. The lympho-
cyte count progressively increased, again with the same
phenotype and reached 35×109/L with associated ane-
mia. The patient was treated with ESHAP, rituximab
and autologous bone marrow transplantation16and
achieved a second durable complete remission which
still persists as of September 2007. Informed consent
was obtained from the patient and the study was
approved by the INSERM review board (October 2005).
Chromosome studies using RHG and GTG banding
techniques were performed on peripheral blood cells.
Description of chromosomal changes followed the
ISCN recommendations.17Metaphase FISH analysis
was performed as previously described.18
Nucleic acid methods
RNA and DNA extraction and analyses were per-
formed using standard protocols. Nylon membranes
were Hybond N+ (Amersham, Les Ulis, France). Inverse-
PCR was performed as previously described19using
inverse primers as follows: external-REV GCAGT-
GAGTGAGAGGTCAGCA, external-FOR GACAGT-
GAATAATGGCCCTACA, internal-REV GGTCCAGT-
CAACTTCCTGCT, and internal-FOR GTTACATTG-
CACATGATGACTATA. PCR products were then
sequenced. Classical PCR was then used to confirm
location and sequence of the breakpoint (FOR
GCAAATCAAGGTGGCAAGGA AND REV GCAGT-
Sequence analyses were performed locally or at
A human NKX2-5 HA-tagged expression construct
was obtained by PCR amplification of the entire open
reading frame, was cloned into the vector pcDNA3, and
verified by DNA sequencing.
For RT-PCR analyses, 2 µg of total RNA extracted
from patients’ mononuclear bone marrow cells or cell
lines were used to synthesize cDNA using the
Superscript II kit (Invitrogen, Cergy-Pontoise, France)
and random hexamers. PCR amplifications were per-
formed starting from 2 µL of cDNA template for 33
cycles consisting of 94°C for 1 min; 55°C for 30 secs.,
and 72°C for 30 secs., primers used were as follow:
NKX2-5-FOR TCTATCCACGTGCCTACAGC; NKX2-
Probes for Southern blot
A probe corresponding to chromosome 5 sequences
upstream of NKX2-5, spanning nucleotides 172596357
to 172598356, was obtained by PCR amplification of
BAC 281H14 DNA using AAAGACACAGCTCC-
CGCAGGC and ACGAAGAGCAGAGTCGCGCT
primers. The J δ1 probe is a 1.7 kb XbaI fragment span-
ning nucleotides 21988328 to 21990065 of chromo-
50 µg of total cellular extracts were separated in 10%
SDS-PAGE gel and then blotted to nitrocellulose mem-
branes (Schleicher & Schuell, Cassel, Germany). 1:1,000
Anti NKX2-5 (SC-14033, Santa Cruz, Palo Alto, CA,
USA) and 1:2,000 anti-actin (A5316, Sigma Aldrich, St-
Quentin-Fallavier, France) antibodies were used.
Results and Discussion
Characterization of the chromosomal translocation
At diagnosis in 1995 (P-1), the karyotype was reported
as 46,XX,t(5;14)(q34;q11)/46, idem, i(8)(q10)/46,
XX (Figure 1A). This translocation was present during
all the disease course. In 2002 (P-2) the karyotype was
9,add(11)(q24),+mar2[cp8]/46,XX. FISH studies with
IGH-CCDN1 probes ruled out a cryptic t(11;14)
(q13;q32). In keeping with the location of the break-
point, FISH analyses demonstrated split of TCR δ
encompassing probes (Figure 1B). The use of FISH
probes aimed at detecting the rearrangement of the
TLX3 locus on chromosome 5 resulting from the com-
mon t(5;14) translocation showed signals on both nor-
mal and rearranged chromosomes 5, indicating that the
breakpoint was telomeric to this locus (data not shown).
To check for involvement in the NKX2-5 gene, we used
BAC 466H21 as a FISH probe. This probe generated a
signal on the der(5) and der(14) chromosomes in addi-
tion to a signal on the normal chromosome 5 (Figure
1B). Together, these data indicate that the TCRD gene
was rearranged within the NKX2-5 locus as a result of
this t(5;14) translocation. To establish involvement of
the TCRD gene at the molecular level, we used a probe
encompassing the J δ 1 segment in Southern blotting
experiments. This probe identified an abnormal frag-
ment in addition to the germline band in DNA digested
with either BamHI or BglII (Figure 2A). The same
NKX2-5 activation in mature B-cell leukemia
haematologica | 2008; 93(7) | 1083 |
rearrangement was observed at two different stages of
the disease (P-1 and P-2). To better characterize the
breakpoint on chromosome 5, we used an inverse PCR
strategy starting from TCR sequences using DNA from
P-1. Analyses of the nucleotide sequence of the ampli-
fied fragment located the breakpoint on chromosome 5,
2.5 kb downstream from the NKX2-5 gene (see Online
Supplementary Figure 1 for a scheme of the loci). To con-
firm our result, we selected a probe from this region for
Southern blot analyses of the patient's DNA. The probe
detected an abnormal fragment upon digestion with
EcoRI in addition to the germline band. This probe also
identified an abnormal BglII fragment of similar size to
the abnormal fragment detected by the J δ 1 probe (data
not shown). To further compare the two stages of disease,
the nucleotide sequence of the breakpoint was deter-
mined after PCR amplification of the patient's DNA and
was shown to be identical. Nucleotide sequence align-
ment of the chromosomal breakpoint with its normal
counterpart is shown in Figure 2C. This allowed us to
locate the breakpoint on chromosome 14 immediately
5' of the D δ 3 segment. The sequence of the chromoso-
Figure 1. Cytogenetic analysis of patient P-1 (A) Partial karyotype of the patient (B) FISH analysis of P-1 metaphase. BAC probes are as
follows RP11-466H21 (NKX2-5, red) and RP11-262M15 (TCRD, green).
5 der (5 14 der (14)
Figure 2. Analyses of the t(5;14) in patient P (A) Southern blot analyses of patient DNA for TCR δ gene status. A probe corresponding
to the Jδ1 segment reacts with an abnormal fragment (indicated by an arrowhead) in patient's material with respect to a wild type con-
trol (N). Note that the same rearranged fragments are detected at both stages of the disease (P-1, P-2). (B) Southern blot analyses of
patient DNA for NKX2-5 gene status. An identical rearrangement (indicated by an arrowhead) is also detected with a chromosome 5
probe at both stages of the disease. (C) Alignment of the t(5;14) fusion sequences and wild type counterparts. N nucleotides appear in
bold lowercase letters. Recombination signals (RS) from Dδ3 segment and cryptic RS sequences from chromosome 5 are underlined.
(D) RT-PCR analyses of NKX2-5. A specific fragment was amplified from cDNA from patient's material and in positive controls. (E)
Western blotting analyses of NKX2-5 in patient sample. The NKX2-5 protein is detected as a doublet in patient material (P-1), as well
as positive controls (IVT and CCRF-CEM). IVT: in vitro translation. Actin was used as a loading control. MOLT4 does not express NKX2-5.
CCRF-CEM and MOLT4 are established human T-ALL cell lines used as positive and negative control respectively.
Probe: J delta 1 Probe: 5’ NKX2-5
N P-2 P-1
N P-2 P-1
X. Su et al.
| 1084 | haematologica | 2008; 93(7)
mal fusion demonstrated the presence of non-templated
(N) nucleotides implicating V(D)J recombinase activity
in the genesis of the translocation. On chromosome 5,
recombination-related sequences, i.e. a canonical 21 bp
spacer between heptamer and nonamer sequences,
were observed which might account for the location of
the chromosome 5 breakpoint.
RT-PCR analysis demonstrated the expression of
NKX2-5 at both stages of the disease, whereas this gene
was not expressed in 41 B-cell lymphoma samples
devoid of t(5;14) (Figure 2D and data not shown). NKX2-
5 protein expression was confirmed in P-1.
Transcriptional activation through TCRαδ
rearrangement in a B-type neoplasm is a rare event.20, 21
However, since the TCR enhancers are known to be
lymphoid specific but not T-cell restricted, the function-
al consequences of these rearrangements are transcrip-
tional activation of the partner genes.
To compare the biological properties of NKX2-5 to
those of TLX1 and HOXA9, we used a myeloid colony
formation assay. Murine hematopoietic precursors
(Linneg/low) were transduced by MSCV based constructs
expressing NKX2-5 and plated in methylcellulose. The
colonies were scored and compared with empty viruses
or viruses encoding HOXA9 or TLX1. All retroviruses
were similarly efficient in transduction (data not shown).
All three constructs conferred enhanced replating poten-
tial to the transduced cells (Figure 3A). The growth stim-
ulatory effect towards myeloid progenitors was more
pronounced for HOXA9 and TLX1, but NKX2-5 expres-
sion repeatedly conferred to myeloid progenitors the
ability to generate colonies after the third round of
replating, a feature never observed with empty vector.
The moderate effect of NKX2-5 markedly differs from
the effect of HOXA9 and is comparable to the effect of
TLX3 or HOXA13.22Homeobox proteins have docu-
mented effects on cell survival, proliferation and fate.
The three orphan homeobox genes, TLX1, TLX3 and
NKX2-5, are known to be transcriptionally activated in
T-ALL. During development, TLX1 is involved in gene-
sis of the spleen, and TLX1 and TLX3 play a role in cel-
lular fate determination of some neuron subtypes.23The
cardiac homeobox protein, NKX2-5, is essential to car-
diac development and mutations of NKX2-5 cause vari-
ous congenital heart diseases.24It would be of interest to
identify and compare the oncogenic pathways triggered
by those oncoproteins. Ectopic expression of NKX2-5
has also been reported in two T-cell lines derived from
T-cell malignant diseases and in a T-ALL patient.12This
report also supports an oncogenic role for NKX2-5. The
frequency of HOX gene activation in malignant B-cell
diseases must still be determined, keeping in mind that
our patient with a t(5;14) translocation exhibited an
unusual extranodal disease.
Given the involvement of the TCR δ gene, the
translocation might have occurred at an early step of
lymphoid differentiation. Nevertheless, our results
demonstrate that homeobox transcriptional activation
occurs in both T- and B-lymphoid proliferation and also
in both acute and chronic disease.
Authorship and Disclosures
XS designed research and perfomed research; VD-V
perfomed research; ED provided vital reagents and
revised the manuscript; ZA provided vital reagents; RB
designed research and revised the manuscript; HM-B
provided vital reagents and revised the manuscript;
OAB funded research and revised the manuscript; FN-K
designed research and wrote the manuscript.
The authors reported no potential conflicts of interest.
1. Boehm T, Rabbitts TH. The human
T cell receptor genes are targets for
chromosomal abnormalities in T cell
tumors. Faseb J 1989;3:2344-59.
2. Weng AP, Ferrando AA, Lee W,
Morris JPt, Silverman LB, Sanchez-
Irizarry C, et al. Activating muta-
tions of NOTCH1 in human T cell
3. Armstrong SA, Golub TR, Kors-
meyer SJ. MLL-rearranged leu-
kemias: insights from gene expres-
sion profiling. Semin Hematol 2003;
4. Aspland SE, Bendall HH, Murre C.
The role of E2A-PBX1 in leukemoge-
nesis. Oncogene 2001;20:5708-17.
5. van Grotel M, Meijerink JP, Beverloo
HB, Langerak AW, Buys-Gladdines
Figure 3 (left). Properties of myeloid precursors transduced with
HOXA9, NKX2-5, TLX1. (A) Clonogenic properties of bone marrow
cells upon serial replating on methylcellulose. Colony numbers
are shown per 5×103cells. Results are average of 4 experiments.
(B) Western blot of cell extracts from primary colonies from MIE
(empty vector), MSCV-HOXA9, MSCV-NKX2.5, MSCV-TLX1-trans-
duced cells. Equal amounts of protein were loaded and blots were
probed with anti NKX2-5. Analyses of GFP expression and FACS
analyses showing the myeloid nature of the transduced cells are
shown in Online Supplementary Figure 2.
round IIround III
Number of colonies
NKX2-5 activation in mature B-cell leukemia
haematologica | 2008; 93(7) | 1085 |
JG, Schneider P, et al. The outcome
of molecular-cytogenetic subgroups
in pediatric T-cell acute lymphoblas-
tic leukemia: a retrospective study of
patients treated according to DCOG
or COALL protocols. Haematologica
6. Ballerini P, Blaise A, Busson-Le
Coniat M, Su XY, Zucman-Rossi J,
Adam M, et al. HOX11L2 expression
defines a clinical subtype of pediatric
T-ALL associated with poor progno-
sis. Blood 2002;100:991-7.
7. Berger R, Dastugue N, Busson M,
Van Den Akker J, Perot C, Ballerini P,
et al. t(5;14)/HOX11L2-positive T-
cell acute lymphoblastic leukemia. A
collaborative study of the Groupe
Francais de Cytogenetique Hemato-
logique (GFCH). Leukemia 2003;17:
8. Hansen-Hagge TE, Schafer M, Kiyoi
H, Morris SW, Whitlock JA, Koch P,
et al. Disruption of the RanBP17/
Hox11L2 region by recombination
with the TCRd locus in acute lym-
phoblastic leukemias with t(5;14)
(q34;q11). Leukemia 2002;16:2205-
9. Bernard OA, Busson-LeConiat M,
Ballerini P, Mauchauffe M, Della
Valle V, Monni R, et al. A new recur-
rent and specific cryptic transloca-
tion, t(5;14)(q35;q32), is associated
with expression of the Hox11L2
gene in T acute lymphoblastic
leukemia. Leukemia 2001;15:1495-
10. Su XY, Della-Valle V, Andre-Schmutz
I, Lemercier C, Radford-Weiss I,
Ballerini P, et al. HOX11L2/TLX3 is
transcriptionally activated through
T-cell regulatory elements down-
stream of BCL11B as a result of the
t(5;14)(q35;q32). Blood 2006;108:
11. Nagel S, Kaufmann M, Drexler HG,
MacLeod RA. The cardiac homeobox
gene NKX2-5 is deregulated by juxta-
position with BCL11B in pediatric T-
ALL cell lines via a novel t(5;14)
(q35.1;q32.2). Cancer Res 2003;63:
12. Przybylski GK, Dik WA, Grabarczyk
P, Wanzeck J, Chudobska P, Jan-
kowski K, et al. The effect of a novel
recombination between the home-
obox gene NKX2-5 and the TRD
locus in T-cell acute lymphoblastic
leukemia on activation of the NKX2-
5 gene. Haematologica 2006;91:317-
13. Matutes E, Owusu-Ankomah K,
Morilla R, Garcia Marco J, Houlihan
A, Que TH, Catovsky D. The immu-
nological profile of B-cell disorders
and proposal of a scoring system for
the diagnosis of CLL. Leukemia
14. Harris NL, Jaffe ES, Diebold J, Flan-
drin G, Muller-Hermelink HK,
Vardiman J, et al. The World Health
Organization classification of neo-
plasms of the hematopoietic and
lymphoid tissues: report of the
Clinical Advisory Committee meet-
ing-Airlie House, Virginia, Novem-
ber, 1997. Hematol J 2000;1:53-66.
15. Leporrier M, Chevret S, Cazin B,
Boudjerra N, Feugier P, Desablens B,
et al. Randomized comparison of flu-
darabine, CAP, and ChOP in 938 pre-
viously untreated stage B and C
patients. Blood 2001;98:2319-25.
16. Sutton L, Maloum K, Gonzalez H,
Zouabi H, Azar N, Boccaccio C, et al.
Autologous hematopoietic stem cell
transplantation as salvage treatment
for advanced B cell chronic lympho-
cytic leukemia. Leukemia 1998;12:
17. ISCN. An International System for
Human Cytogenetic Nomeclature.
LG Shaffer and N Tommerup (ed); S.
Karger, Basel, 2005.
18. Romana SP, Cherif D, Le Coniat M,
Derre J, Flexor MA, Berger R. In situ
hybridization to interphase nuclei in
acute leukemia. Genes Chromo-
somes Cancer 1993;8:98-103.
19. Su XY, Busson M, Della Valle V,
Ballerini P, Dastugue N, Talmant P, et
al. Various types of rearrangements
target TLX3 locus in T-cell acute
lymphoblastic leukemia. Genes
Chromosomes Cancer 2004;41:243-
20. Park JK, McKeithan TW, Le Beau
MM, Bitter MA, Franklin WA,
Rowley JD, et al. An (8;14)(q24;q11)
translocation involving the T-cell
receptor alpha-chain gene and the
MYC oncogene 3' region in a B-cell
lymphoma. Genes Chromosomes
21. Duro D, Bernard O, Della Valle V,
Leblanc T, Berger R, Larsen CJ.
Inactivation of the P16INK4/MTS1
gene by a chromosome translocation
t(9;14)(p21-22;q11) in an acute lym-
phoblastic leukemia of B-cell type.
Cancer Res 1996;56:848-54.
22. Su X, Drabkin H, Clappier E, Mor-
gado E, Busson M, Romana S, et al.
Transforming potential of the T-cell
acute lymphoblastic leukemia-asso-
ciated homeobox genes HOXA13,
TLX1, and TLX3. Genes Chromo-
somes Cancer 2006;45:846-55.
23. Cheng L, Arata A, Mizuguchi R,
Qian Y, Karunaratne A, Gray PA, et
al. Tlx3 and Tlx1 are post-mitotic
selector genes determining gluta-
matergic over GABAergic cell fates.
Nat Neurosci 2004;7:510-7.
24. Akazawa H, Komuro I. Too much
Csx/Nkx2-5 is as bad as too little? J
Mol Cell Cardiol 2003;35:227-9.