Nuclear Factor Y Drives Basal Transcription of the Human TLX3, a Gene Overexpressed in T-Cell Acute Lymphocytic Leukemia

Silvia Borghini,1 Manuela Vargiolu,1 Marco Di Duca,2 Roberto Ravazzolo,1,3 and Isabella Ceccherini1

1Laboratorio di Genetica Molecolare and 2Laboratorio di Fisiopatologia dell’Uremia, Istituto Giannina Gaslini; 3Dipartimento di Pediatria e CEBR, Universita`di Genova, Genova, Italy

Abstract Introduction Based on a knocked-out mouse model and a few TLX3 (also known as HOX11L2 or Rnx) belongs to the TLX expression studies, TLX3 is regarded as a homeobox family of homeodomain transcription factors that includes two gene crucial for the development of the autonomic other members: TLX1 (also known as HOX11) and TLX2 (also nervous system. This gene can undergo rearrangements known as HOX11L1, Ncx,orEnx). Members of this gene or deregulation, giving rise to T-cell acute lymphocytic family exhibit restricted, and often overlapping, patterns of leukemia. The present report is focused on the expression, including developing neural crest-derived tissues, identification of elements and factors playing a role in as shown by in situ hybridization studies (1-4). TLX3 is the TLX3 physiologic expression regulation and uniquely expressed in the developing medulla oblongata and is therefore likely to be involved in cancer development. required for proper formation of first-order relay visceral In particular, after identifying the transcription start sensory neurons and of most of the (nor)adrenergic centers points, we have made use of in vitro transfection assays in the brainstem, especially involved in the physiologic control to show that the 5¶-untranslated region of the gene is of cardiovascular and respiratory systems (5). Accordingly, necessary for the basal promoter activity in cell lines TLX3-deficient mice displayed hypoventilation, sudden apneas, from different origin. By site-directed mutagenesis, and cyanosis, and all died within 24 hours after birth from a two tandem CCAAT boxes have been localized as hypothesized central defect in respiratory control, thus having critical elements of this region. In vivo chromatin been regarded as a model for congenital central hypoventilation immunoprecipitation and electrophoretic mobility shift syndrome (6), but extensive screening of this gene failed to assays have indicated that nuclear factor Y (NFY) detect any mutation in patients (7, 8). recognizes these sites in all the analyzed cell lines. Expression of TLX3 has also been detected in leukemia The physiologic role of such an interaction has been samples from 20% of children and 13% of adults affected with confirmed by a dominant-negative version of the NFY T-cell acute lymphocytic leukemia (9-12), although this gene transcription factor that has turned out to decrease both has never been involved in normal T-cell differentiation (13). in vitro TLX3 promoter activity and endogenous The TLX3 ectopic expression has been associated mainly with amount of mRNA. Finally, a consistent in vivo TLX3 the occurrence of chromosomal translocations involving 5q35 expression impairment was also achieved after NFY (14, 15), where the gene is located (16, 17), but it is still unclear mRNA knockdown. The full characterization of the TLX3 whether the underlying molecular pathogenetic mechanism transcription regulation will ultimately provide crucial relies either in the loss of negative regulatory elements on elements to define the involvement of this gene in chromosome 5q or in the juxtaposition of positive transcrip- T-cell acute lymphocytic leukemia development. tional regulatory elements of genes lying on the partner (Mol Cancer Res 2006;4(9):635–43) chromosomes. To date, no detailed information is available regarding the biology of the TLX3 gene and cellular pathways in which it is involved, although based on its specific spatiotemporal expression and the correlation between up- regulation and development of T-cell acute lymphocytic Received 11/27/05; revised 6/23/06; accepted 7/5/06. leukemia, a role of this gene in either cell cycle, survival, or Grant support: Compagnia di San Paolo and Italian Telethon grant GGP04257 differentiation can be postulated. (I. Ceccherini). As a first step to identify factors and DNA elements crucial The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in for the transcriptional regulation of the TLX3 gene and able to accordance with 18 U.S.C. Section 1734 solely to indicate this fact. account for both neuronal and leukemia development, we have Note: Present address for M. Vargiolu: U.O. Genetica Medica, Policlinico S. focused on its 5¶-regulatory region. In particular, we have Orsola-Malpighi, Bologna, Italy. Requests for reprints: Isabella Ceccherini, Laboratorio di Genetica Molecolare, showed that the 5¶-untranslated region (UTR) of TLX3 contains Istituto Giannina Gaslini, L.go Gerolamo Gaslini, 5, 16148 Genova, Italy. Phone: the most important cis-acting elements required for its 39-10-5636800; Fax: 39-10-3779797. E-mail: [email protected]. Copyright D 2006 American Association for Cancer Research. expression and identified the transcription factor nuclear factor doi:10.1158/1541-7786.MCR-05-0250 Y (NFY) as the main regulator binding this sequence.

Results regarded as major transcription start point. According to the Determination of Transcriptional Start Sites notion that the presence of multiple transcription start points To identify the transcription start site(s) of TLX3, we took is a common feature of TATA-less promoters (22), three advantage of a protocol for the automated detection of additional peaks of minor intensity were detected (Fig. 1, fluorescent primer extension products, adapted from the peaks 1, 3, and 4, respectively). Interestingly, the major standard method based on radioactive labeling (18, 19) and transcription start point, as well as the start site 1, is located already successfully applied to other genes (20, 21). upstream of two CCAAT boxes, whereas start sites Initial experiments were done to set up the system using 3 and 4 map between them. variable amounts of RNA (2.5, 5, and 10 Ag) extracted from We could confirm these results in additional experiments IMR32 cells, which express TLX3, and two different done using RNA from SHSY-5Y neuroblastoma cells, which concentrations of fluorescent primer (1or 10 Amol/L). We express TLX3, and RNA obtained from a second extraction run the primer extension reactions together with a molecular from IMR32 cells (Fig. 1). weight marker, and this allowed to observe that signals obtained from different primer extension products were always Comparison between Mouse and Human TLX3 Gene in the same position, with an intensity proportional to the Sequences amount of RNA used (data not shown). With 10 Amol/L To characterize crucial elements involved in the transcrip- primer concentration, nonspecific background fluorescence tion regulation of the TLX3 gene, we run computer-assisted occurred in the form of an increased baseline, thus suggesting analysis finalized to identify evolutionary conserved sequen- 1 Amol/L as the most appropriate concentration for subsequent ces and putative binding sites for specific factors using experiments. To obtain a precise determination of the mVISTA (http://pipeline.lbl.gov/cgi-bin/gateway2) and Mat- transcriptional start point(s), we also run the primer extension Inspector version 2.2 software (23), respectively. As shown in reactions together with the DNA fragments generated in a Fig. 2, by comparing 5 kb of human and mouse DNA sequencing reaction carried out with the same but unlabeled sequences corresponding to the complete TLX3 gene and its oligonucleotide (20). The peak detected 272 nucleotides 5¶-flanking region, we observed an overall degree of upstream of the ‘‘ATG’’ translation start codon (Fig. 1, conservation of >75% in the 1,300 bp located immediately peak 2) showed the highest intensity and was therefore upstream of the ‘‘ATG’’ translation start codon. In particular,

FIGURE 1. Determination of the transcriptional start sites of the TLX3 gene. Top box, DNA sequence obtained with the unlabeled primer; middle and bot- tom boxes, chromatograms of two cell lines (IMR32 and SH- SY5Y, respectively) obtained loadinginthesamewellthe sequencing reaction and the primer extension product carried out with the 6-FAM primer. The fluorograms are reversed to match the sequence of the coding strand shown in Fig. 2. Transcription start points can be recognized as additional peaks corresponding to the C nucleotide (complementary to the G nucleotide revealed by 6-FAM fluorescence) laying on the original sequence and marked by arrows. Based on the intensity, peak 2 has been considered the major transcription start point.

FIGURE 2. Sequence conservation analysis between human and mouse TLX3 genes. Y axis, percentage of sequence identity detected between human and mouse using the mVISTA software; X axis, reference DNA seg- ment. Black bar arrow above the plot, TLX3 gene; blue rectangles, exons; light blue rectangles, UTRs. SINE and other repetitive elements are also indicated. Con- served regions in which every contiguous subsegment of 100-bp length was at least 60% identical to its paired sequence are high- lighted in blue, pink, and light blue for coding, noncoding, and UTR regions, respectively. Boxbelow the plot, 1,311 bp of the 5¶-region of human TLX3 selected for the study. Arrows, distal limit of each of the deleted constructs used in the transfection assays. The CCAAT repetitions are boxed. Positions are numbered from the major transcription start point of the gene (+1; red arrow).

360 bp, including the 5¶-UTR, showed an identity of 98.3%. contribution of the cis-acting elements identified by computer- In this region, MatInspector pointed out the lack of assisted analysis to the promoter activity, we transfected six conventional TATA boxes and the presence of two tandem deletion constructs obtained by removing DNA fragments of identical CCAAT boxes. increasing lengths at the distal end of the region under investigation. As shown in Fig. 3, only the deletion mapping Functional Characterization of the Promoter Activityof the at +98 affected transcription, leading to statistically significant Human TLX3 Gene loss of detectable activity of the reporter gene, whereas the We have cloned 1,311 bp of the 5¶-region adjacent to the removal of shorter fragments had no negative effect. Actually, ATG of the human TLX3 gene mentioned above into the the sequence +32/+98 was sufficient to guarantee a level of pGL3-basic vector, upstream of the reporter firefly luciferase luciferase expression comparable with that obtained trans- gene, as a first step to study the promoter activity by transient fecting the whole 1,311-bp 5¶-region in all the tested cell lines cellular transfections. Because TLX3 is specifically expressed (Fig. 3, top construct), thus suggesting that crucial elements in neural crest-derived tissues, we used as cell recipients two for the transcriptional activation of TLX3 localize in this neuroblastoma cell lines that express the gene, SHSY-5Y and proximal sequence, a region showing a very high level of IMR32, and two cell lines from different lineages that do not conservation during evolution, as already observed (Fig. 2). express the gene, the adenocarcinoma MCF-7 and the The sequence +32/+98 is characterized by the presence of two osteosarcoma SaOS-2. The construct displayed well-detectable ‘‘CCAAT’’ boxes whose site-directed mutagenesis reduced the transcriptional activity in all the tested cell lines, ranging from promoter activity to <50% of that displayed by the wild-type 25% in MCF-7 to 80% in SaOS-2 relative to a control (WT) 1,311-bp sequence in all the analyzed cell lines (Fig. 4). plasmid containing the SV40 promoter, thus confirming that A further reduction was observed when both sites were the fragment under analysis contains regulatory sequence of mutated, suggesting that transcription factors binding these the human TLX3 gene (data not shown). To establish the sites do have an additive effect.

FIGURE 3. Functional analysis of the 5¶-regulatory sequence of human TLX3. Transient transfections of constructs containing progressively deleted fragments of the promoter fused to firefly luciferase gene were carried out in four different cell lines. Y axis, nucleotide position, relative to the major start point, of the 5¶-end of each construct; X axis, relative luciferase activity corrected for luciferase activity of the promoter-less pGL3 vector. Asterisk, the only deletion construct that induced, in all the four cell lines under analysis, a statistically significant decrease in the reporter gene expression levels with respect to the full-length reporter-promoter construct (À1,039 bp). In particular, the following probabilities (P) were esti- mated: 0.00023 (SHSY-5Y), 0.019 (IMR32), 0.022 (SaOS-2), and 0.00021 (MCF-7).

Interestingly, the two CCAAT boxes, responsible for a basal NFY effectively immunoprecipitated the TLX3 regulatory transcriptional activity in TLX3-expressing and nonexpressing sequence under analysis (À95/+189) from all these cell lines. cells, are located in the 5¶-UTR, a gene region that therefore Control assays, done in SHSY-5Y cells both without antibody seems necessary for inducing TLX3 expression. and in the presence of different unrelated antibodies (a-p53, a- E2F-1, a-h-actin, and normal mouse IgG), and in MCF-7 cells NFY Binding the +32/+98 Sequence using a-h-actin did not show any significant DNA amplified Matrix sequences of the CCAAT boxes in the proximal from the same region. Finally, an amplification product was 5¶-regulatory region of the TLX3 gene were indicated as obtained also from SHSY-5Y and MCF-7 nucleoproteins peculiar of NFY binding following MatInspector analysis. To precipitated by an a-acetyl-histone H4 antibody, thus suggest- investigate the possible in vivo interaction between NFY and ing that this region is characterized by an open configuration the above sequence, we did chromatin immunoprecipitation chromatin even in cells that do not express TLX3 (Fig. 5). assays using formaldehyde cross-linked and sonicated chroma- To clarify whether NFY binds both CCAAT boxes, tin from SHSY-5Y, MCF-7, and IMR32 cells. PCR results electrophoretic mobility shift assays (EMSA) were done using (Fig. 5) showed that an antibody specific for the A subunit of nuclear extracts prepared from IMR32 cells. When two

FIGURE 4. Site-directed mutagenesis of the CCAAT boxes. A. Transcriptional luciferase activity driven by mutated constructs, relative to the activity of the WT construct (100%), with numerical values. Decreases in the reporter gene expression due to presence of single or double mutations were all highly statistically significant. B. Sequence diversity of the three mutant constructs tested. In particular, the distal CCAAT box1 site is mutated in MUT1, the proximal CCAAT box 2 site is mutated in MUT2, and both CCAAT boxes are mutated in MUT3.

oligonucleotides spanning regions corresponding to either CCAAT boxes 1and 2 were incubated with IMR32 nuclear extract, one major complex was detected in both cases (Fig. 6, lanes 1 and 5). The sequence-specific nature of this interaction was confirmed through abolition of the complex by a 200-fold molar excess of unlabeled WT competitor (Fig. 6, lanes 4 and 6). Oligonucleotides bearing the mutations tested in transfection experiments had no effect in the analysis of FIGURE 5. In vivo interaction between NFY and 5¶-regulatory region of CCAAT box 1(Fig. 6, lane 3) but displayed an effect on the TLX3 gene. Chromatin immunoprecipitation assays done on chromatin CCAAT box 2 although weaker than the WT competitor (Fig. 6, derived from SHSY-5Y, IMR32, and MCF-7 cells. Primers specific for À95/ lane 7). However, in both cases, preincubation with an antibody +189 sequence were used to amplify DNA from complexes immunoprecipitated with an a-NFYA subunit antibody (a-NFY). INPUT, fragmented specific for NFY subunit A induced the supershift of the DNA before immunoprecipitation; NoAb, no antibody immunoprecipitation retarded bands, showing the involvement of NFY in binding the control; a-acH4, antibody against hyperacetylated histone H4; a-act, h two CCAAT boxes (Fig. 6, lanes 2, 8, and 9). The specificity of antibody against -actin. such interactions is well proven by the observation that the supershifted bands were competed by a 200-fold molar excess a specific small interfering RNA pool. As shown in Fig. 8, these of the unlabeled WT oligonucleotides (data not shown for small interfering RNAs, transfected into SHSY-5Y, hindered the CCAAT box 1; Fig. 6, lane 10 for CCAAT box 2). Because synthesis of the targeted transcription factor determining a the mutant CCAAT box 2 oligonucleotide partially competed specific effect on both mRNA (Fig. 8A) and protein (Fig. 8B). with the probe (Fig. 6, lane 11), to confirm the specificity of the As expected, when the NFYA expression was reduced, the level interaction between NFY and the CCAAT box 2 sequence, we of TLX3 expression was significantly diminished with respect incubated the reaction mixture with an unrelated oligonucleo- to that obtained after the transfection of a control small tide containing the consensus binding site for Sp1(Fig. 6, interfering RNA pool. lane 12). Differently from what expected in nonspecific interaction between DNA, proteins, and antibody, the unrelated Discussion Sp1oligonucleotide was not able to destabilize this complex, TLX3 is a homeobox gene that plays a fundamental role confirming the NFY-specific recognition of the CCAAT box 2. during the development of nervous tissues derived from neural crest cells. Deregulated expression of this gene is also detected Effect of a Dominant-Negative Version of NFYA in several patients affected with T-cell acute lymphocytic We did experiments in which the complete 5¶-regulatory leukemia. Thus far, nothing is known about the regulation of region under analysis (À1,039LUC) or the identified minimal TLX3 expression both under physiologic conditions, during sequence (+32LUC) was cotransfected in SHSY-5Y neuroblas- embryogenesis, and in pathology, during leukemia onset. toma cells and MCF-7 adenocarcinoma cells together with an In the current study, we have focused on mechanism(s) expression plasmid containing DNFYA13m19, which is regulating transcription of the human TLX3 gene and showed a characterized by a NFY dominant-negative activity already pivotal role of the 5¶-UTR. A similar location of crucial binding described (24). As shown in Fig. 7A, overexpression of sites for transcription factors has already been reported for other DNFYA13m29 resulted in marked reduction of luciferase genes as exemplified by the human plasminogen gene (25), activity induced by both the long and the short promoter the A g-globin gene (26), and the human monocyte chemo- constructs with respect to control cotransfections done using the attractant protein-1receptor gene (27). DNFYA13 WT construct. The most critical regulatory site of the human TLX3 5¶-UTR Moreover, we investigated the effect of the same dominant- seems to be recognized by the transcription factor NFY, a negative version of NFY on the endogenous level of TLX3 complex composed of three subunits (NFYA, NFYB, and mRNA. After transfection of SHSY-5Y cells with both NFYC) that are highly conserved through evolution among DNFYA13 expression plasmids, comparable levels of mutant eukaryotes (28) and all necessary for DNA binding (29). NFY and WT NFY proteins, clearly higher than those associated with binds CCAAT motifs in regulatory regions of many constitu- the empty vector transfection, were induced (Fig. 7B). tively active genes and, in spite of its ubiquitous pattern of Subsequently, RNA extraction, cDNA synthesis, and real-time expression, of cell type–specific genes. In addition, NFYactivity PCR allowed to reveal a relative level of TLX3 mRNA in the has been shown to be modulated during differentiation (30, 31), dominant-negative DNFYA13m29-transfected cells dramatical- proliferation (32), and senescence (33). Although NFY recog- ly lower than that obtained with the empty vector (Fig. 7C). On nition sites are usually located at a relatively fixed position in the the other hand, as expected, because NFY works as a trimer, promoter, between 80 and 100 bp upstream of the transcription the DNFYA13 WT construct does not have any effect on the start site (28), an increasing number of exceptions has been TLX3 transcription. Taken together, these observations confirm described (34); in particular, NFY binding 5¶-UTR has already the functional role played by NFY in the transcription regulation been reported for proper regulation of human CD34 gene (35). of TLX3. EMSA experiments have shown that NFY specifically binds two tandem CCAAT boxes in the 5¶-UTR of the TLX3 gene. Effect of NFYA Knockdown Both site-directed mutagenesis of these sites and forced To definitely confirm the in vivo role of NFY in TLX3 gene expression of a dominant-negative version of NFY had a regulation, we carried out the knockdown of NFY subunit A by dramatic effect on the activity of the luciferase reporter gene

driven by the 5¶-regulatory region of TLX3. These in vitro results were confirmed in vivo at the endogenous TLX3 gene level. In particular, NFY binding the 5¶-UTR of the TLX3 gene was unequivocally shown by chromatin immunoprecipitation analysis in three cell lines, both neuronal and not neuronal, and NFY-dependent TLX3 gene transcription was strengthened by down-regulation of its mRNA under two different conditions: forced expression of a dominant-negative NFYA and endogenous NFYA knockdown. NFY has been shown to exert a cell type–specific activity of gene expression regulation (30). However, after observing comparable luciferase activities in cell lines from different origin in our transfection experiments, the tandem CCAAT boxes can be regarded as unique proximal cis-acting elements sufficient for NFY-mediated transcription of TLX3 in all the analyzed cell lines. NFY might constitutively bind the two CCAAT boxes, located in the TLX3 5¶-UTR, with a major role of either attracting the basal transcription machinery or, as a result of its association with histone acetyltransferases (36), maintaining the chromatin structure in an open configuration. Results from our chromatin immunoprecipitation assays suggest that this is the case also in TLX3 nonexpressing cells (MCF-7), thus confirming the constitutive nature of the NFY binding to the identified regulatory region. Distal elements, possibly favoring the access of repressively acting factors to the promoter, would then account for the TLX3-specific expression during embryogenesis. Such a hypothesis could also explain the ectopic TLX3 expression in leukemia, which would be due to FIGURE 7. Effect of NFY dominant-negative protein on the TLX3 gene transactivation. A. Luciferase activity of either À1040LUC or +31LUC dislocation of the gene and of its basal constitutively active sequences was measured on lysates from SHSY-5Y and MCF-7 cells promoter with respect to those cis-elements normally driving its cotransfected with expression plasmid containing WT (DNFYA13) or dominant-negative NFYA (DNFYA13m29). Fold induction versus activity cell-specific transcription. Long-range regulation is increasingly derived from transfection of pGL3-basic vector alone. *, P V 0.05, significant levels of difference between WT and mutant NFY constructs. B. Western blot obtained using 4 Ag lysates from cells transfected with either the empty expression vector, the expression plasmid containing WT (DNFYA13), or the dominant-negative NFYA (DNFYA13m29). a-act, antibody specific for h-actin. C. Endogenous levels of TLX3 mRNA were measured by real-time PCR on cDNA obtained from SHSY-5Y cells transfected with the above plasmids. Columns, TLX3 transcript quantity relative to that of mock-transfected cells. Only the mutant construct has resulted to significantly decrease the TLX3 transcript (*, P < 0.01).

recognized as a means by which developmentally significant transcription control occurs (37, 38). In addition, a recent report has provided a functional mapping of such a kind of regulation for TLX1 (39), a gene belonging to the same family and that, similarly to TLX3, is abnormally expressed in a subset of T-cell acute lymphocytic leukemia patients (40). Therefore, for both genes, a mechanism driving their ectopic expression can be postulated which would rely in either the loss of distal negative regulation elements on the chromosome regions where they are located or the juxtaposition of positive transcriptional regulatory elements of genes lying on the other involved partner chromosomes. The failure to detect gross cytogenetic abnormalities in some T-cell acute lymphocytic leukemia patients characterized by the expression of TLX1 or TLX3 has suggested that an alternative activation mechanism may exist (10, 41), FIGURE 6. In vitro interaction between NFY and CCAAT boxes. EMSA assays were done incubating nuclear extracts from IMR32 cells with biotin- thus enforcing the hypothesis of deletion or mutation of labeled double-stranded oligonucleotide containing either CCAAT box1 or dominant distal negative cis-acting regulators. 2. WT, WT competitor 200Â; MUT, mutated competitor 200Â; Sp1, competitor 200Â representing the Sp1 consensus binding sequence; Ab, In conclusion, we have reported robust evidence on how the antibody a-NFYA; SS, supershifted band; asterisk, major complex. basal expression of the human TLX3 gene is achieved. Further

investigations will be needed to clarify the mechanism(s) both extension reactions were loaded, in a same gel well, together with conferring cell and developmental stage specificity to TLX3 either a molecular weight marker (ROX 500, AB) or the DNA expression and eventually predisposing to its involvement in fragments generated in a sequencing reaction carried out with the leukemia. same but unlabeled oligonucleotide. Runs of the extended products were done in an ABI PRISM 3130 DNA sequencer Materials and Methods (Applied Biosystems, Foster City, CA) and data thus generated Primer Extension Analysis were analyzed by using GeneScan and Sequence Analysis, Total RNA was extracted from IMR32 and SHSY-5Y respectively. A neuroblastoma cells by both Trizol reagent (Invitrogen, Carls- The experiment was repeated using 10 g RNA from bad, CA) and RNeasy columns (Qiagen, Hilden, Germany). both IMR32 (a different extraction) and SHSY-5Y cells with A Primer 5¶-GAGGCGGGCAGAGACGGGTATGTGGCG-3¶, 1 mol/L final concentration of primer. designed downstream of the ATG first codon, was labeled at the 5¶-terminus with the fluorochrome 6-FAM. Total RNA (2.5, Reporter Gene Constructs 5, or 10 Ag) from IMR32 cells was hybridized with the above Reporter constructs were prepared using the pGL3-basic vector (Promega, Madison, WI), which contains firefly primer (1or 10 Amol/L final concentration) in H2O by heating at 70jC for 5 minutes and subsequently incubating at 30jC for luciferase as reporter gene. Cloning of the 1,311-bp sequence 15 minutes. For the extension reactions, annealing solutions lying upstream of the ATG first codon of TLX3 was done as were combined with 800 units Moloney murine leukemia virus described previously (7). Progressive deletions of the promoter reverse transcriptase and its specific buffer (Clontech, Mountain sequence were obtained by exonuclease III digestion using the View, CA), deoxynucleotide triphosphates each at 5 mmol/L Erase-a-Base System according to the manufacturer’s instruc- final concentration, 40 units RNase inhibitor (BD Clontech), and tions (Promega). 150 ng/AL actinomycin D. cDNA syntheses were run at 42jC for Site-directed mutagenesis has been achieved by PCR (42) 2 hours and stopped by heating 5 minutes at 95jC. EDTA (1 AL; using the WT construct as template to obtain single mutated 0.5 mmol/L, pH 8.0) supplemented with 5 ng RNase A/DNase constructs from which, finally, a construct with both mutated free was then added for 30 minutes at 37jC to remove RNA, CAAT boxes has been generated. potentially affecting the electrophorectic mobility of the Oligonucleotides used for site-directed mutagenesis were synthesized cDNA. Reactions were phenol extracted and each Mut 1(+31/54)sense 5 ¶-AGCCGGGAGCCTCTGAGAGGA- ¶ ¶ TT was subdivided into four aliquots before ethanol precipitation. AGAG-3 and Mut 2 (+63/86) sense 5 -TGATCGCAGCC - ¶ Samples were air dried and redissolved in formamide. Primer TGGCTGCGGCA-3 . Mutated nucleotides are represented in bold. Each construct was controlled by direct sequencing, and MatInspector analysis (23) was run both to check the loss of the specific putative binding site and to exclude the appearance of any additional different site.

Assayfor the Promoter Activity A total of 105 cells were plated on 35-mm Petri dishes 1day before transfection. Constructs under analysis and Renilla luciferase reporter plasmid pRL-CMV (Promega), used as a transfection efficiency control, were cotransfected using Fugene 6 (Roche, Basel, Switzerland). The promoter-less pGL3-basic vector and the pGL3-promoter vector, containing the SV40 promoter, were used as negative and positive controls, respectively. Constructs (90 fmol) under analysis were mixed with 10 fmol pRL-CMV and added to 1 AL Fugene 6. In cotransfection experiments, 1 Ag expression vectors containing DNFYA13 WT or DNFYA13m29 dominant negative (kind gifts of Roberto Mantovani, Dipartimento di Scienze Biomolecolari e Biotecnologie, Universita`di Milano, Italy; ref. 24) was added to the DNA solution. For the IMR32 neuroblastoma cell line, constructs under analysis (375 fmol) and Renilla luciferase reporter plasmid pRL-CMV (150 fmol) were cotransfected using polyethylenimine method. After transfection, cells were incubated for 2 hours at 37jC and polyethylenimine solution was then FIGURE 8. NFYA knockdown. A. NFYA and TLX3 transcript levels in substituted with fresh complete medium. SHSY-5Y cells were normalized to the transcript level of b2-microglobulin and expressed as fold to the mock-transfected SHSY-5Y. Columns, mean After 24 hours (48 hours for cotransfections), transfected of four experiments each done in triplicate; bars, SD. *, P < 0.05; **, cells were lysed with PLB 1Â buffer (Promega). Luciferase P <0.01(t test). B. Western blot of NFYA protein level in 4 Ag lysates from cells transfected with either siNFYA or siControl. Filter was reprobed with activities in cell lysates were determined using Promega a-h-actin antibody as control of total lysates quantity. protocol (Dual Luciferase Reporter Assay System) and

luminometer (Turner Designs, Sunnyvale, CA). Each construct Analysis of Endogenous TLX3 mRNA was transfected in duplicate and each test was repeated at least SHSY-5Y neuroblastoma cells (1 Â 106) were transfected thrice. by Fugene 6 with 6 Ag expression vectors containing DNFYA13 WT or DNFYA13m29 dominant negative (24). Electrophoretic MobilityShift Assay The same quantity of empty expression vector was trans- Nuclear extracts were prepared as described previously (43) fected as negative control. Forty-eight hours later, total from IMR32 cells using the following solutions: solution A RNA was extracted using RNeasy Mini kit and DNase (10 mmol/L HEPES, 1.5 mmol/L MgCl2, 10 mmol/L KCl, treatment (Qiagen) following manufacturer’s instructions. 0.5 mmol/L DTT, 0.5 mmol/L phenylmethylsulfonyl fluoride), Spectrophotometer-quantified RNA (500 ng) from each BLS solution (20 mmol/L HEPES, 1.5 mmol/L MgCl2,0.5 sample was retrotranscribed using Advantage RT-for- mmol/L DTT, 0.5 mmol/L phenylmethylsulfonyl fluoride, 0.2 PCR kit (BD Clontech). Real-time quantitative PCR was mmol/L EDTA, 20 mmol/L NaCl), and BHS solution carried out using inventoried Assays-on-Demand provided (20 mmol/L HEPES, 1.5 mmol/L MgCl2, 0.5 mmol/L DTT, by Applied Biosystems (Hs.00253271_m1 for TLX3 and 0.5 mmol/L phenylmethylsulfonyl fluoride, 0.2 mmol/L EDTA, Hs.99999907_m1for b2-microglobulin, used as endogenous 0.9 mol/L NaCl). EMSA assays were done using LightShift reference) using an ABI Prism 7700 Sequence Detection chemiluminescent EMSA kit (Pierce, Rockford, IL) following System (Applied Biosystems). After having ascertained that the manufacturer’s instructions. Briefly, binding reactions were efficiencies of TLX3 and b2-microglobulin are approximately carried out for 20 minutes at room temperature in binding buffer equal, changes in mRNA amount of TLX3 were quantified 1Â containing 50 ng/AL poly(deoxyinosinic-deoxycytidylic by using the comparative CT Method (Sequence Detection acid), 50 mmol/L KCl, and 5% glycerol using 20 fmol biotin- System Chemistry Guide, Applied Biosystems). Real- end-labeled target DNA and 10 Ag nuclear extracts. For time PCR amplification was done in triplicate and repeated competition and supershift assays, unlabeled target DNA thrice. (4 pmol) or 1 AL anti-NFYA subunit antibody (Santa Cruz Biotechnology, Santa Cruz, CA) were added, respectively. Small Interfering RNA–Based NFYA Knockdown DNA-protein complexes were run in nondenaturing polyacryl- SHSY-5Y neuroblastoma cells were transfected in 60-mm amide gels, transferred onto a positively charged nylon Petri dishes using 15 AL LipofectAMINE 2000 (Invitrogen, membrane (Hybond-N+), UV cross-linked, and finally detected Carlsbad, CA) and 500 pmol siNFYA pool or siControl pool using horseradish peroxidase–conjugated streptavidin. (Dharmacon, Lafayette, CO) according to the manufacturer’s Double-stranded oligonucleotides used in EMSA experiments instructions. After growing in antibiotic-free medium for ¶ were CCAAT box 1sense 5 -AGCCGGGAGCCAATGAGAG- 48 hours, total RNA was extracted by Qiagen Mini columns, ¶ ¶ GAGAG-3 and CCAAT box 2 sense 5 -TGATCGCAGCCAA- retrotranscribed, and analyzed by quantitative real-time reverse ¶ TGGCTGCGGCA-3 . Mutated versions are the same used in transcription-PCR as described above (Assay-on-Demand used transfection assays. for NFYA Hs.00242929_m1).

Chromatin Immunoprecipitation Assay Western Blot Chromatin immunoprecipitation assays were done on SHSY- SHSY-5Y cells were transfected with the NFY expression 5Y, IMR32, and MCF-7 cells following the protocol provided by the manufacturer (Upstate Biotechnology, Lake Placid, NY). constructs or small interfering RNAs as described above. Briefly, after cross-linking obtained by incubation with Forty-eight hours later, cells were lysed and the total amount A formaldehyde at 1% final concentration for 10 minutes at of proteins was quantified. Each sample (4 g) were resolved 37jC and subsequent quenching with 125 mmol/L glycine, on an SDS-PAGE, transferred onto polyvinylidene difluoride h cells were lysed in SDS buffer in the presence of the protease membrane, and hybridized with a-NFYA or a- -actin and inhibitors cocktail (Roche) and 0.5 mmol/L phenylmethylsul- finally with rabbit a-mouse IgG-horseradish peroxidase. fonyl fluoride. Samples were sonicated and supernatants Detection was achieved using chemiluminescence reagent containing fragmented chromatin were precleared by adding enhanced chemiluminescence plus (Amersham, Munich, salmon sperm-DNA protein A-agarose beads. A little portion of Germany). the supernatants was kept as ‘‘input’’ material. The remaining cleared chromatin was incubated overnight with 5 Agofthe Statistical Analysis following antibodies: anti-NFYA subunit (Santa Cruz Biotech- Statistical differences in the levels of reporter gene nology) to study NFY-DNA interaction, anti-acetyl-histone H4 expression driven by each progressively deleted portion of (Upstate Biotechnology) to study chromatin condensation state, the TLX3 5¶-flanking region, with respect to an entire sequence normal mouse IgG (Upstate Biotechnology), anti-p53 (Santa of 1,311 bp, were determined by applying the Student’s t test. Cruz Biotechnology), anti-E2F-1(Santa Cruz Biotechnology), The same test was also used to compare means of different and anti-h-actin (Sigma, St. Louis, MO), of the same isotype experimental points. In particular, statistical significance was of the anti-NFY as negative controls. Finally, an overnight assessed for (a) the preventing effect of the mutated CCAAT incubation with no antibody was also done. Genomic sequences boxes in driving reporter gene expression, (b)thein vitro and of interest were amplified by PCR using primers forward in vivo effects of the NFY dominant-negative construct, and (À95/À77) 5¶-TCGGGCTCTAATATCCGCC-3¶ and reverse (c)theTLX3 expression decrease induced by the NFY (+189/+171) 5¶-CCTCTCGTCGCACTGAAAC-3¶. knockdown.

Acknowledgments 21. Patek M, Muth G, Wohlleben W. Function of Corynebacterium glutamicum We thank Dr. Roberto Mantovani for the kind gift of expression plasmids promoters in Escherichia coli, Streptomyces lividans, and Bacillus subtilis. containing DNFYA13 WT and DNFYA13m29 dominant negative, Dr. Fabio J Biotechnol 2003;104:325 – 34. Pastorino for the h-actin antibody, and Loredana Velo for excellent secretarial 22. Azizkhan JC, Jensen DE, Pierce AJ, Wade M. Transcription from TATA-less assistance. promoters: dihydrofolate reductase as a model. Crit Rev Eukaryot Gene Expr 1993;4:229 – 54. References 23. Quandt K, Frech K, Karas H, Wingender E, Werner T. MatInd and 1. Hatano M, Iitsuka Y, Yamamoto H, et al. Ncx, a Hox11 related gene, is MatInspector—new fast and versatile tools for detection of consensus matches in expressed in a variety of tissues derived from neural crest cells. Anat Embryol nucleotide sequence data. Nucleic Acids Res 1995;23:4878 – 84. (Berl) 1997;195:419 – 25. 24. Mantovani R, Li XY, Pessara U, Hooft van Huisjduijnen R, Benoist C, 2. Logan C, Wingate RJ, McKay IJ, Lumsden A. Tlx-1and Tlx-3 homeobox Mathis D. Dominant negative analogs of NF-YA. J Biol Chem 1994;269: gene expression in cranial sensory ganglia and hindbrain of the chick embryo: 20340 – 6. markers of patterned connectivity. J Neurosci 1998;18:5389 – 402. 25. Meroni G, Buraggi G, Mantovani R, Taramelli R. Motifs resembling 3. Patterson KD, Krieg PA. Hox11-family genes XHox11 and XHox11L2 in hepatocyte nuclear factor 1and activator protein 3 mediate the tissue specificity of Xenopus: XHox11L2 expression is restricted to a subset of the primary sensory the human plasminogen gene. Eur J Biochem 1996;236:373 – 82. neurons. Dev Dyn 1999;214:34 – 43. 26. Amrolia PJ, Cunningham JM, Ney P, Nienhuis AW, Jane SM. Identification ¶ 4. Uchiyama K, Otsuka R, Hanaoka K. CHox11L2, a Hox11 related gene, is of two novel regulatory elements within the 5 -untranslated region of the human A expressed in the peripheral nervous system and subpopulation of the spinal cord g-globin gene. J Biol Chem 1995;270:12892 – 8. during chick development. Neurosci Lett 1999;273:97 – 100. 27. Yamamoto K, Takeshima H, Hamada K, et al. Cloning and functional ¶ 5. Qian Y, Fritzsch B, Shirasawa S, Chen CL, Choi Y, Ma Q. Formation of characterization of the 5 -flanking region of the human monocyte chemoattractant ¶ brainstem (nor)adrenergic centers and first-order relay visceral sensory neurons is protein-1receptor (CCR2) gene. Essential role of 5 -untranslated region in tissue- dependent on homeodomain protein Rnx/Tlx3. Genes Dev 2001;15:2533 – 45. specific expression. Biol Chem 1999;274:4646 – 54. 6. Shirasawa S, Arata A, Onimaru H, et al. Rnx deficiency results in congenital 28. Mantovani R. A survey of 178 NF-Y binding CCAAT boxes. Nucleic Acids central hypoventilation. Nat Genet 2000;24:287 – 90. Res 1998;26:1135 – 43. 7. Matera I, Bachetti T, Cinti R, et al. Mutational analysis of the RNX gene 29. Maity SN, Sinha S, Ruteshouser EC, de Crombrugghe B. Three different in congenital central hypoventilation syndrome. Am J Med Genet 2002;113: polypeptides are necessary for DNA binding of the mammalian heteromeric 178 – 82. CCAAT binding factor. J Biol Chem 1992;267:16574 – 80. 8. Amiel J, Pelet A, Trang H, et al. Exclusion of RNX as a major gene in 30. Marziali G, Perrotti E, Ilari R, et al. The activity of the CCAAT-box binding congenital central hypoventilation syndrome (CCHS, Ondine’s curse). Am J Med factor NF-Y is modulated through the regulated expression of its A subunit during Genet 2003;117:18 – 20. monocyte to macrophage differentiation: regulation of tissue-specific genes through a ubiquitous transcription factor. Blood 1999;93:519 – 26. 9. Ballerini P, Blaise A, Busson-Le Coniat M, et al. HOX11L2 expression defines a clinical subtype of pediatric T-ALL associated with poor prognosis. 31. Gurtner A, Manni I, Fuschi P, et al. Requirement for down-regulation of the Blood 2002;100:991 – 7. CCAAT-binding activity of the NF-Y transcription factor during skeletal muscle differentiation. Mol Biol Cell 2003;14:2706 – 15. 10. Mauvieux L, Leymarie V, Helias C, et al. High incidence of Hox11L2 expression in children with T-ALL. Leukemia 2002;16:2417 – 22. 32. Bhattacharya A, Deng JM, Zhang Z, Behringer R, de Crombrugghe B, Maity SN. The B subunit of the CCAAT box binding transcription factor complex (CBF/ 11. Berger R, Dastugue N, Busson M, et al. t(5;14)/HOX11L2-positive T-cell NF-Y) is essential for early mouse development and cell proliferation. Cancer Res acute lymphoblastic leukemia. A collaborative study of the Groupe Francais de 2003;63:8167 – 72. Cytogenetique Hematologique (GFCH). Leukemia 2003;17:1851 – 7. 33. Matuoka K, Chen KY. Transcriptional regulation of cellular ageing by the 12. Cave H, Suciu S, Preudhomme C, et al. Clinical significance of HOX11L2 CCAAT box-binding factor CBF/NF-Y. Ageing Res Rev 2002;4:639 – 51. expression linked to t(5;14)(q35;q32), of HOX11 expression, and of SIL-TAL fusion in childhood T-cell malignancies: results of EORTC studies 58881and 34. Testa A, Donati G, Yan P, et al. Chromatin immunoprecipitation (ChIP) on 58951. Blood 2004;103:442 – 50. chip experiments uncover a widespread distribution of NF-Y binding CCAAT sites outside of core promoters. J Biol Chem 2005;280:13606 – 15. 13. Ferrando A, Herblot S, Palomero T, et al. Biallelic transcriptional activation of oncogenic transcription factors in T-cell acute lymphoblastic leukemia. Blood 35. Radomska HS, Satterthwaite AB, Taranenko N, Narravula S, Krause DS, 2004;103:1909 – 11. Tenen DG. A nuclear factor Y (NFY) site positively regulates the human CD34 stem cell gene. Blood 1999;94:3772 – 80. 14. Bernard OA, Busson-LeConiat M, Ballerini P, et al. A new recurrent and specific cryptic translocation, t(5;14)(q35;q32), is associated with expression of 36. Currie RA. NF-Y is associated with the histone acetyltransferases GCN5 and the Hox11L2 gene in T acute lymphoblastic leukemia. Leukemia 2001;15: P/CAF. J Biol Chem 1998;273:1430 – 4. 1495 – 504. 37. Townes TM, Behringer RR. Human globin locus activation region (LAR): 15. Su XY, Busson M, Della Valle V, et al. Various types of rearrangements target role in temporal control. Trends Genet 1990;7:219 – 23. TLX3 locus in T-cell acute lymphoblastic leukemia. Genes Chromosomes Cancer 38. Lettice LA, Hill RE. Preaxial polydactyly: a model for defective long- 2004;41:243 – 9. range regulation in congenital abnormalities. Curr Opin Genet Dev 2005;15: 16. Cinti R, Fava M, Sancandi M, Matera I, Ravazzolo R, Ceccherini I. 294 – 300. Assignment of the HOX11L2 gene to human chromosome band 5q35.1 and of its 39. Brake RL, Chatterjee PK, Kees UR, Watt PM. The functional mapping of murine homolog to mouse chromosome bands 11A4-5 by in situ hybridization. long-range transcription control elements of the HOX11 proto-oncogene. Cytogenet Cell Genet 2001;92:354 – 5. Biochem Biophys Res Commun 2004;313:327 – 35. 17. Lee-Kirsch MA, Engel K, Paditz E, Rosen-Wolff A, Lee YA, Gahr M. 40. Dube ID, Kamel-Reid S, Yuan CC, et al. A novel human homeobox gene lies Assignment of the human homeobox 11-like 2 gene (HOX11L2) to chromosome at the chromosome 10 breakpoint in lymphoid neoplasias with chromosomal 5q34->q35 by radiation hybrid mapping. Cytogenet Cell Genet 2001;93:358. translocation t(10;14). Blood 1991;78:2996 – 3003. 18. Yamada M, Izu H, Nitta T, Kurihara K, Sakurai T. High-temperature 41. Kees UR, Heerema NA, Kumar R, et al. Expression of HOX11 in childhood nonradioactive primer extension assay for determination of transcription-initiation T-lineage acute lymphoblastic leukaemia can occur in the absence of cytogenetic site. BioTechniques 1998;25:72 – 8. aberration at 10q24: a study from the Children’s Cancer Group (CCG). Leukemia 19. Altermann E, Klein JR, Henrich B. Synthesis and automated detection 2003;5:887 – 93. of fluorescently labeled primer extension products. BioTechniques 1999;26: 42. Higuchi R, Krummel B, Saiki, RK. A general method of in vitro preparation 96 – 101. and specific mutagenesis of DNA fragments: study of protein and DNA 20. Blasco JL, Garcı´a-Sa´nchez MA, Ruiz-Herrera J, Eslava AP, Iturriaga EA. A interactions. Nucleic Acids Res 1988;16:7351 – 67. gene coding for ornithine decarboxylase (odcA) is differentially expressed during 43. Dignam JD, Lebovitz RM, Roeder RG. Accurate transcription initiation by the Mucor circinelloides yeast-to-hypha transition. Res Microbiol 2002;153: RNA polymerase II in a soluble extract from isolated mammalian nuclei. Nucleic 155 – 64. Acids Res 1983;11:1475 – 89.

Mol Cancer Res 2006;4(9). September 2006 Downloaded from mcr.aacrjournals.org on September 26, 2021. © 2006 American Association for Cancer Research. Nuclear Factor Y Drives Basal Transcription of the Human TLX3, a Gene Overexpressed in T-Cell Acute Lymphocytic Leukemia

Silvia Borghini, Manuela Vargiolu, Marco Di Duca, et al.

Mol Cancer Res 2006;4:635-643.

Updated version Access the most recent version of this article at: http://mcr.aacrjournals.org/content/4/9/635

Cited articles This article cites 40 articles, 15 of which you can access for free at: http://mcr.aacrjournals.org/content/4/9/635.full#ref-list-1

Citing articles This article has been cited by 2 HighWire-hosted articles. Access the articles at: http://mcr.aacrjournals.org/content/4/9/635.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mcr.aacrjournals.org/content/4/9/635. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.