|Home | About | Journals | Submit | Contact Us | Français|
Telomerase is an RNA-dependent DNA polymerase that synthesizes telomeric DNA. Its activity is not detectable in most somatic cells but it is reactivated during tumorigenesis. In most cancers, the combination of hTERT hypermethylation and hypomethylation of a short promoter region is permissive for low-level hTERT transcription. Activated and malignant lymphocytes express high telomerase activity, through a mechanism that seems methylationin-dependent. The aim of this study was to determine which mechanism is involved in the enhanced expression of hTERT in lymphoid cells. Our data confirm that in B cells, some T cell lymphomas and non-neoplastic lymph nodes, the hTERT promoter is unmethylated. Binding sites for the B cell-specific transcription factor PAX5 were identified downstream of the ATG translational start site through EMSA and ChIP experiments. ChIP assays indicated that the transcriptional activation of hTERT by PAX5 does not involve repression of CTCF binding. In a B cell lymphoma cell line, siRNA-induced knockdown of PAX5 expression repressed hTERT transcription. Moreover, ectopic expression of PAX5 in a telomerase-negative normal fibroblast cell line was found to be sufficient to activate hTERT expression. These data show that activation of hTERT in telomerase-positive B cells is due to a methylation-independent mechanism in which PAX5 plays an important role.
The telomerase enzyme allows germ cells, stems cells and cancer cells to divide indefinitely . Human telomerase possesses a highly regulated subunit called telomerase reverse transcriptase (hTERT), which is the limiting factor for its activity [2,3]. The hTERT expression is nearly imperceptible in the majority of differentiated somatic cells, which lead to inevitable telomeric attrition and subsequently cellular senescence. High levels of hTERT are detected in proliferative somatic cells such as endometrial tissues or activated lymphocytes, but also in most immortalized and cancer cells.
hTERT transcription has been shown to be influenced by numerous activators and inhibitors, such as c-Myc, Sp1, Hif-1, Mbi-1, USF1/2, oestrogen response element, p53, Mad1, myeloid-specific zinc finger protein 2 (MZF-2), TGFβ, Wilms’ tumour 1 (WT1) and CTCF [4–6]. In addition, a possible role of DNA methylation in hTERT transcription regulation can be expected, as the hTERT promoter is situated within a dense CG-rich CpG island. In normal somatic cells the hTERT promoter is unmethylated, although the transcription of the gene is repressed. However, in most cancer cells, hypermethylation of this region correlates with expression of the gene and with perceptible telomerase activity [7–10]. This apparent contradiction with the classical mechanism of transcriptional repression by DNA methylation was recently clarified. DNA methylation exhibits a dual role in hTERT transcriptional regulation by interfering with the binding of inhibitors, such as the CTCF transcription factor, and by partial hypomethylation of the core promoter, which allows the hTERT gene to be permissive for transcription . Furthermore, the absence of methylation in association with active chromatin marks around the transcription start site of hTERT indicates that expression and DNA methylation patterns of the hTERT promoter are not in contradiction to the general model of gene silencing mediated by DNA methylation .
In a small subset of telomerase-positive tumours, hTERT expression appears to be regulated by a methylation-independent mechanism [12–14]. For example, the hTERT promoter is methylated in only 30% of ovarian cancers, almost all of which are telomerase-positive . Cells of the lymphoid system also seem to escape methylation-dependent mechanism of hTERT regulation. Leukaemias and lymphomas, including B cell chronic lymphocytic leukaemia (CLL), express high levels of telomerase but exhibit low levels of hTERT promoter methylation . More recently, acute myeloid leukaemia (HL-60) and Burkitt’s lymphoma (Raji) cell lines, as well as normal lymphocytes, were found to have hypomethylated hTERT promoters .
B cells exhibit notably longer telomeres than any other blood cell population, such as T cells, natural killer cells and monocytes . B cells in the germinal centre of tonsils have longer telomeres compared to naïve and memory B cells . As expected, low levels of telomerase activity are observed in naïve and memory B cells, in contrast to germinal centre B cells that exhibit high telomerase activity [16,17].
Paired box (PAX) proteins include nine members that are important regulators in early development for tissue specificity . Once bound to DNA, PAX proteins can play the role of transcriptional activators or repressors [19–21]. Deregulation of PAX genes has been associated with a variety of cancers, including astrocytoma, medulloblastoma, lymphoma and Wilm’s tumour [22,23]. Moreover, PAX expression has been suggested to be essential for survival of cancer cells. Recently, PAX8 has been implicated in the activation of hTERT and hTR promoters, which in turn activate telomerase in glioma . PAX2, PAX5 and PAX8 belong to the same subgroup and thus could impact on hTERT regulation in a tissue-specific manner. During B cell development, the PAX5 gene is expressed in early B cell precursors (pro-B cells) and continues to be expressed up to mature B cells, but not in terminally differentiated plasma cells [25,26]. As a consequence, PAX5 expression is used as a lineage-specific marker in B cells neoplasms [27,28]. PAX5 has been shown to promote the expression of target genes encoding crucial components of the (pre)BCR signalling cascade, such as the receptor signalling chain Igα, also called CD79a and mb-1 [29,30], the costimulatory receptor CD19 [21,31] and the central adaptor protein BLNK . PAX5 also facilitates the VH–DJH recombination step and can activate other transcription factor genes .
Our working hypothesis for the experiments reported in this paper was that hTERT regulation in B cells is methylation-independent. To confirm our hypothesis, methylation status of the hTERT promoter was investigated in normal and malignant lymphoid tissues. Then, the B cell-specific factor PAX5 was considered for its participation in the induction of hTERT expression in telomerase-positive B cells.
The Burkitt’s lymphoma cell lines Daudi and Ramos and the pre-B cell leukaemia line Nalm6 were kindly provided by Dr Benedicte Baisse (CHUV, Lausanne, Switzerland). The Burkitt line, Raji, was kindly provided by Apoxis (Lausanne, Switzerland). HeLa (cervical adenocarcinoma), PC-3 (prostate adenocarcinoma) and BJ (normal fibroblasts) cells were obtained from ATCC (Manassas, VA, USA). Cell lines were cultured in the medium recommended by ATCC.
The lymphomas and normal tissues came from the files of the Institute of Pathology of Lausanne. The samples included: four histologically non-neoplastic lymphoid tissues (two lymph node biopsies and two tonsil biopsies); six B cell non-Hodgkin lymphomas (NHLs), comprising three high-grade and three low-grade NHLs; and six T cell NHLs. All diagnoses were confirmed by a pathologist (RB). The lymphoma cases were selected to ensure that sufficient populations of T or B cells were present. The use of human tissues for this study was done according to the guidelines of the local ethics committee.
Total RNA of frozen tissues and cultured cells was extracted using Trizol-LS (Invitrogen, Basel, Switzerland) according to the manufacturer’s protocol. The extraction protocol for fixed tissues was described previously . RT–PCRs were performed using Super- Script One-Step RT–PCR or Quantitative RT–PCR ThermoScript™ One-Step System (Invitrogen) (for primers and RT–PCR for each individual gene, see Supporting information, Table S1).
DNA was extracted from frozen and fixed tissues and cultured cells using the DNeasy tissue kit (Qiagen). DNA (2 µg) was modified with sodium bisulphite and used to amplify a 224 bp fragment of the hTERT promoter, as previously described . PCR products were analysed by a methylation-sensitive dotblot assay (MS-DBA)  and confirmed by direct sequencing  and methylation-sensitive single-strand conformation analysis (MS-SSCA) [36,37].
Oligonucleotides of the hTERT exonic region and the CD79A promoter region (hTERT, sense, 5′-GCTGGTGCAGCGCGGGGACCCGGCGGCTTT-3′; CD79A, sense 5′-AGCGAGGGCCACTGGAGCCCATCTCCGGGG-3′) were labelled with the DIG-Oligonucleotide 3′-End Labelling Kit (Roche). Gel shift reactions were performed using the DIG Gel Shift Kit (Roche) with 0.5 pmol DIG-labelled oligonucleotide and 5 µg Nalm6 or Raji cell extracts. A supershift assay was performed with a PAX5 rabbit antibody (Active Motif, Carlsbad, USA) on Raji cell extracts, according to the manufacturer’s protocol.
ChIP assays were performed using EZ ChIP (Upstate Biotechnology, Lake Placid, NY, USA), following the manufacturer’s instructions with some modifications. After crosslinking with 1% formaldehyde and sonication to shear DNA, lysates from 2 × 106 cells (PAX5) or 4 × 106 cells (CTCF) were diluted in ChIP dilution buffer for immunoprecipitation or stored at 4 °C to be directly uncrosslinked and purified (DNA input fraction). Magnetic beads (40 µl, Dynabeads Protein G, Invitrogen) were incubated for 1 h at room temperature in 60 µl of the blocking solution with either 2 µg goat polyclonal anti-PAX5 antibody (Santa Cruz, CA, USA), 10 µg mouse polyclonal anti-CTCF antibody (Rockville, MD, USA) or without antibody. After washing and incubation overnight at 4 °C with the chromatin solution, the beads were washed twice with the following solutions: low-salt, high-salt, LiCl and finally Tris-EDTA (TE). The eluate was then resuspended in 200 µl 5% Chelex solution and incubated for 10 min at 100 °C to reverse the protein–DNA crosslinks. After purification, the immunoprecipitated DNA was analysed by quantitative real-time PCR with specific primers (see Supporting information, Table S2). The human CTCF-binding site N, a MYC insulator site (MYC-N), and H19 were used as positive controls and a CTCF non-binding site, G of MYC (MYC-G) was used as a negative control [38,39]. For chromatin immunoprecipitation of PAX5, CD19, which is a well-known target of PAX5, was used as a positive control, whereas KRAS, which does not contain PAX5 binding sites, was used as a negative control.
Antigen retrieval was performed using a pressure cooker for 2 min in 10 mm sodium citrate buffer, pH 6. The slides were incubated overnight at 4 °C with the anti-TERT antibody (1 : 50, EST21-A; Alpha Diagnostic International, San Antonio, TX, USA), the anti-PAX5 monoclonal antibody (1 : 50, BD Biosciences Pharmingen, San Jose, CA, USA) or the CD3 monoclonal antibody (1 : 1, Novocastra, Newcastle upon Tyne, UK). After washing, the EnVision+ System – HRP AEC (TERT) or the EnVision+ Peroxidase rabbit followed by DAB staining (PAX5 and CD3) were used according to the manufacturer’s instructions (Dako, Glostrup, Denmark). The slides were then counterstained with haematoxylin.
Jet PEI transfection reagent (2 µl) (Polyplus-transfection, Illkirch, France) was used to transfect, in 105 normal fibroblast BJ cells, 2 µg PAX5 expression plasmid (phPAX5, a kind gift from Professor M Busslinger, Research Institute of Molecular Pathology, Vienna, Austria) . Cells treated the same way but without plasmid were used as a transfection control. Dnase extraction and total RNA extraction were performed 48 h after transfection.
A double-stranded annealed Stealth RNAi oligonucleotide targeting PAX5 was designed by Invitrogen software (sense, 5′-GAGGAUAGUGGAACUUGCUCAUCAA-3′). A non-specific fluorescent siRNA (Invitrogen) was used as a control. Transfection of 130 pm siRNA oligonucleotides in 4 × 106 Raji cells was performed with Amaxa Nucleofector (Amaxa Biosystems, Cologne, Germany) according to the manufacturer’s protocol. The efficiency of RNA silencing was checked by western blot with PAX5 antibody (BD Biosciences, Erembodegem, Belgium) and confirmed by quantitative RT–PCR.
Quantitative RT–PCRs were performed on a Rotorgene 6000 cycler (Corbett Research, Sydney, Australia). hTERT and PAX5 mRNAs were amplified using the Quantitative RT–PCR Thermoscript One-Step System (Invitrogen) (the primers and probes are described in the Supporting information, Table S3). CD19 and β-actin were amplified by the same enzymes, but with 1.25 µm SYTO 9 fluorescent dye (Invitrogen) instead of the labelled probes. The relative level of each mRNA was calculated on the basis of the two standard curve relative quantification method. Gene expressions were normalized to β-actin and to the cells transfected with the non-coding siRNA. At least two independent determinations of fold differences were used to calculate the average fold difference values and associated standard deviations (SDs).
To define the methylation status of the hTERT promoter in lymphoid tissues, we analysed six primary B cell lymphomas, six primary T cell lymphomas and four non-malignant lymphoid tissues. Four human lymphoid tumour cell lines were also investigated. RT–PCR analysis confirmed that hTERT transcripts were present in all the lymphoma tissues and cell lines as well as in the non-neoplastic lymphoid tissues (Figure 1A). The β-actin gene was simultaneously amplified as a control.
Using MS-DBA, we next explored the methylation status of the hTERT promoter. In B cell lymphomas and non-neoplastic lymphoid tissues, the hTERT promoter was unmethylated, while it was hypermethylated in half of the T cell lymphomas (Figure 1B). The hTERT promoter was methylated in Daudi cells but unmethylated in the other three cell lines (Raji, Ramos and Nalm6). Direct sequencing and MS-SSCA confirmed the results obtained by MS-DBA (data not shown). To summarize, in some T cell lymphomas hTERT expression goes along with hTERT promoter methylation, as is the case for most solid tumours. In transformed B cells and non-neoplastic lymphocytes, however, hTERT is expressed in the presence of a hypomethylated promoter. In this situation, hTERT expression must be regulated by a methylation-independent mechanism.
To determine whether transcription factors specific to lymphoid cells might be involved in hTERT regulation, we searched for new transcription factor binding sites using MatInspector software (http://www.genomatix.de/matinspector.html). MatInspector revealed two potential binding sites for PAX5, also known as B cell-specific activator protein (BSAP), a transcription factor involved in B cell differentiation and function [25,26], from +110 to +137 bp and +489 to +516 bp downstream of the ATG translational start site (Figure 2A). PAX5 binding sites match the consensus sequence at 9 and 11 out of 15 positions, for exon 1 and exon 2, respectively (Figure 2B). This suggested that PAX5 might be involved in the regulation of hTERT transcription in lymphoid cells.
To determine whether the predicted PAX5 binding sites in the hTERT exon were authentic, we first performed EMSAs using extracts from Raji cells. A specific band for PAX5 was obtained with the CD79A oligonucleotide (Figure 3A, lane 1), which served as a positive control [21,41]. A similar band was obtained with the hTERT probe (lane 2). To check the specificity of the band, cold competitor oligonucleotides were added to the labelled hTERT probe. A 100-fold molar excess of CD79A and hTERT competitors resulted in almost complete inhibition of PAX5 binding (lanes 4 and 6). The same results were obtained with Nalm6 extracts (data not shown). A 5–150-fold increase in the amount of cold CD79A probe also resulted in a progressive inhibition of binding (Figure 3B). Competitive EMSA, with an oligonucleotide in which four specific bases were mutated, did not eliminate PAX5 binding (Figure 3A, lane 5), indicating that PAX5 binding was specific. Pre-incubation of Raji cell extracts with a PAX5-specific antibody resulted in a supershifted band (Figure 3A, lane 7), confirming that PAX5 does bind to the predicted target sequence in the first exon of hTERT.
To determine whether PAX5 bound to the hTERT gene under physiological conditions, chromatin immunoprecipitation (ChIP) experiments were performed, using Nalm6 and Raji cells. DNA samples isolated from the input, the anti-PAX5-bound and the no-antibody fractions were analysed by quantitative real-time PCR. CD19 is a well-known target of PAX5  and was used as positive control. The results indicated an approximately 10-fold enrichment of hTERT exon 1 when normalized with the negative control KRAS gene, and around eight-fold enrichment of hTERT exon 2 (Figure 3D). Enrichment of the CD19 gene was about three to four times greater than that for hTERT, which was not surprising, as the binding of PAX5 to the CD19 target sequences is very strong. In the no-antibody fraction, no enrichment was detected with either hTERT or CD19 (data not shown). Thus, PAX5 is bound in vivo to the first and the second exons of hTERT, providing strong evidence that PAX5 could be involved in the transcriptional regulation of the gene in B cells.
A possible explanation for the effect of PAX5 on hTERT expression could be that it interferes with CTCF-binding to the hTERT promoter. The CTCF transcription factor was found to be essential for repression of hTERT transcription in a variety of normal somatic cells . As PAX5 binding sites lie downstream CTCF target sequences (Figure 2), we therefore performed ChIP analysis to analyse CTCF binding. In Raji cells, hTERT exon 1 was enriched approximately four-fold compared to the negative control, which is in the same range as in the two positive controls, MYC-N and H19 (Figure 4). After transfection with a PAX5 siRNA, a strong reduction in the binding of PAX5 was observed on hTERT and CD19 (Figure 4B), whereas chromatin immunoprecipitation of CTCF did not reveal any significant change in the binding of CTCF to hTERT (Figure 4A). Thus, PAX5 binding to the hTERT exonic region does not block CTCF binding.
PAX5 is a specific marker for all stages of B cell differentiation except for plasma cells . In our series, PAX5 mRNA was detected by RT–PCR in all B and T cell NHLs, as well as in non-neoplastic lymphoid tissues and cell lines. The presence of PAX5 mRNA in the T cell lymphomas could be due to the presence of normal B cells in the tumour tissues.
By IHC of consecutive sections, hTERT and PAX5 were both detected in the same regions of the six B cell lymphomas, suggesting that they were present in the same tumour cells (Figure 5A, B). The T cells were identified by CD3 expression. In B cell lymphomas, the normal T cells did not appear to express either PAX5 or hTERT (Figure 5A, C). In the four non-neoplastic lymphoid tissues, both PAX5 and hTERT were expressed in germinal centre B cells and B cells of the mantle zone (Figure 5G, H), whereas CD3+ T cells were PAX5- and hTERT-negative (Figure 5G–I). The expression of hTERT observed in germinal centre and mantle zone was concordant with the published data on telomerase activity . In the six investigated T cell lymphomas, hTERT was expressed in the CD3+ neoplastic T cells (Figure 5D, F) while PAX5 was not (Figure 5E, F). As expected, in T cell lymphomas, PAX5 was only expressed in normal B cells and therefore PAX5 had no role in activating hTERT expression in tumour T cells. In summary, PAX5 and hTERT co-localize in normal and malignant B cells, supporting the suggestion that PAX5 might be involved in hTERT activation in these cells.
To determine whether a reduction in PAX5 expression would be associated with a change in hTERT expression, we transfected Raji cells with a PAX5 siRNA. After transfection, the down-regulation of PAX5 protein was confirmed by western blot, while the levels of PAX5, hTERT and CD19 transcripts were monitored by quantitative RT–PCR. Cells transfected with a scrambled siRNA were used as a control. Twentyfour hours after transfection, PAX5 transcript levels were reduced ~50% in association with significant reductions in the levels of CD19 and hTERT expression (~30%) (Figure 6A). After 48 h, PAX5 transcripts were reduced by 77% in association with reductions of 57% and 64% in transcripts for CD19 and hTERT, respectively. These studies showed that inhibition of PAX5 leads to a strong down-regulation of hTERT expression, indicating that PAX5 is essential for hTERT expression in B cells.
To determine whether ectopic expression of PAX5 could activate hTERT expression in normal telomerase-negative cells, we transfected normal BJ fibroblasts with a PAX5 expression plasmid. RT–PCR and western blot analyses of PAX5 expression 40 h post-transfection confirmed the siRNA silencing efficiency (Figure 6A). RT–PCR analyses of transcripts for CD19, an established target of PAX5, showed substantial expression in the transfected cells. These data are in agreement with previous studies demonstrating that ectopic expression of PAX5 led the up-regulation to CD19 and other PAX5-target genes [21,43]. Interestingly, the transfected cells expressed hTERT transcripts at similar levels as CD19. The level of expression of hTERT and CD19 was apparently lower in BJ than in Raji cells, this could be explained by the absence in transfected BJ cells of specific transcription factors necessary for a high level of expression of these genes. Non-transfected and mock-transfected BJ cells did not express transcripts of PAX5, CD19 or hTERT. This experiment showed that ectopic expression of PAX5 is sufficient to activate hTERT transcription in normal somatic cells.
In the present study, we showed that hTERT is transcribed in association with the unmethylated 5′ region in B cells, B cell lymphomas and B cell lymphoma cell lines, defining a novel methylation-independent mode of hTERT regulation. EMSA and ChIP assays identified two binding sites in hTERT for the B cell-specific transcription factor PAX5. These sites lie downstream of the ATG translational start site and are located in the first exon and at the beginning of the second exon of hTERT. Moreover, in B cells, decreasing PAX5 expression resulted in a significant reduction in hTERT expression. Importantly, we showed that ectopic expression of PAX5 in telomerase-negative normal cell lines is sufficient to activate hTERT expression. Taken together, these data strongly support a role for PAX5 in the transcriptional activation of hTERT in B cells. Validation of our observations by functional studies will be important.
hTERT is a new PAX5 target, which has no direct link to B cell differentiation, in contrast to the well-known PAX5 target genes. Among the principal targets of PAX5, three genes, CD79A, CD19 and PDCD1 (PD-1), code for cell surface molecules involved in signal transduction, while the products of two other target genes, MYCN (N-Myc) and LEF1, are nuclear transcription factors [31,40]. About 170 PAX5-activated genes have been identified . These genes mediate diverse biological functions in B cells, such as adhesion, migration, signalling and germinal-centre B cell formation, and demonstrate the pleiotropic role of PAX5 in control of the B-lineage commitment.
PAX8, which belongs to the same subgroup of PAX proteins as PAX5, has been implicated in the activation of hTERT in glioma . PAX8 failed to activate the hTERT promoter in telomerase-negative primary cell lines, and other factors seem to be necessary for the expression of hTERT. In contrast, activation of PAX5 was sufficient to initiate the transcription of hTERT in telomerase-negative primary cell lines. Apparently, the action of PAX5 on hTERT is very different from that of PAX8. PAX8 mainly seems to act on the formation of the transcription complex, whereas the major role of PAX5 in transcriptional activation does not seem to be to recruit basal transcription machinery, but is likely to modulate the structure of local chromatin, allowing other sequence-specific factors to activate transcription. Indeed, PAX5 can activate transcription through association with chromatin effector enzymes such as DAXX, CREB-binding protein (CBP) and GCN5, which possess histone acetyltransferase (HAT) activity. PAX5 can also interact with BRG1, a catalytic component of the Swi/Snf chromatin remodelling complexes . On the other hand, CTCF directly binds to SIN3A, which condenses chromatin and prevents transcription by recruitment of histone deacetylase (HDAC) activity . Therefore, the simultaneous binding of CTCF and PAX5 on hTERT exons might produce opposing effects on chromatin: the recruitment of histone modification and nucleosome remodelling activities by PAX5 might antagonize chromatin-mediated transcriptional repression by CTCF. Additional studies need to be performed to more accurately understand how CTCF and PAX5 interact in regulating hTERT expression.
In summary, we describe a methylation-independent mechanism of hTERT regulation that occurs in telomerase-positive B cells. In these cells, hTERT is a novel target of PAX5, which is essential for B cell development and function. According to our data, in B cells, PAX5 also participates in cellular mechanisms underlying cell immortality by up-regulating hTERT gene expression.
This work was supported by grants from the Swiss National Science Foundation (Grants 3100A0-101732 and 3100A0-113505) and in part by the Intramural Research Program of the NIH, National Institute of Allergy and Infectious Diseases.
No conflicts of interest were declared.
SUPPORTING INFORMATION ON THE INTERNET
The following supporting information may be found in the online version of this article:
Table S1. Primer sequences for RT–PCR
Table S2. Primer sequences for ChIP experiments
Table S3. Primer sequences for quantitative RT–PCR