PAX 8 Regulates Telomerase Reverse Transcriptase and Telomerase RNA Component in Glioma

Paired box (PAX) developmental genes are frequently expressed in cancers and confer survival advantages on cancer cells. We have previously found that PAX genes are deregulated in glioma. We have now investigated the expression of PAX genes in glioma and their role in telomere maintenance. The mRNA level of PAX8 showed a positive correlation with telomerase activity in glioma biopsies (r = 0.75, P < 0.001) and in established glioma cell lines (r = 0.97, P = 0.0025). We found that PAX8 is able to coordinately transactivate the promoter for both the telomerase catalytic subunit (hTERT) and the telomerase RNA component (hTR) genes. By electrophoretic mobility shift assay, quantitative PCR, and a telomerase activity assay, we show that PAX8 binds directly to the hTERT and hTR promoters, up-regulating hTERT and hTR mRNA, as well as telomerase activity. Additionally, PAX8 small interfering RNA down-regulated hTERT and hTR . Collectively, these results show that PAX8 may have a role in telomerase regulation. [Cancer Res 2008; 68(14):5724–32]


Introduction
The termini of linear eukaryotic chromosomes are comprised of telomeric repetitive, noncoding DNA, made up of (TTAGGG)n hexanucleotide repeats and associated proteins (1).Telomere ends cannot be fully replicated by the conventional DNA replication machinery, resulting in telomere shortening at each cell division (2).Telomere loss to a critical length can lead to decreased cell viability (3), and increased end-to-end chromosome fusions, which promote genomic instability (4).Thus, loss of terminal nucleotides with each cell division limits the replicative capacity of eukaryotic cells.Therefore, it is believed that cancer cells activate telomere maintenance mechanisms (TMM) to solve this problem and are thus able to replicate indefinitely (5).The most thoroughly studied TMM is telomerase, which is a reverse transcriptase that adds telomeric repeats to chromosomal ends (6).Telomere length may also be maintained in cancer cells by telomerase-independent mechanisms referred to as alternative lengthening of telomeres (7).
Because telomerase can be activated by exogenous TERT expression in cells, TERT has been regarded as a limiting factor in telomerase activation (9).Several lines of evidence, however, indicate that hTR may be limiting (10,11).Both hTERT and hTR are required for a functional telomerase holoenzyme.Telomerase function is stringently regulated at multiple levels, including transcriptional and translational regulation, posttranslational modification, and protein subcellular localization (12).The deregulation of hTERT in cancer cells can be achieved by gene amplification (13).However, most studies have focused on the transcriptional regulation of hTERT expression.Several transcription factors have been shown to transcriptionally regulate hTERT, including c-Myc and Sp1 (14).In addition, multiple tumor suppressor pathways, such as the p53 and transforming growth factor-h pathways, negatively regulate telomerase (15).The hTERT promoter activity is also regulated by hormones and growth factors such as estrogen, basic fibroblast growth factor, and epidermal growth factor (16).
Paired domain-containing transcription factors (PAX) are encoded by a developmental gene family, classified into four subgroups according to the presence or absence of an octapeptide region and the presence, absence, or truncation of a homeodomain (17).PAX genes of the same subgroup are also expressed in similar patterns during development (17).PAX2, PAX5, and PAX8 belong to the same subgroup containing an octapeptide domain and a truncated homeodomain.They are expressed in the midbrainhindbrain junction during brain development (18).The deregulation of PAX genes has been linked with many types of cancer, such as astrocytoma, medulloblastoma, lymphoma, and Wilms' tumor (19,20).In another study, we examined PAX2, PAX5, and PAX8 expression in a panel of 54 gliomas.We found PAX8 to be overexpressed in the majority of these gliomas, whereas PAX2 and PAX5 were overexpressed to a lesser degree. 3Because we previously found TMM to be of prognostic value in glioma (21), we considered the hypothesis that PAX genes, in particular PAX8, might be associated with telomere maintenance, by possibly facilitating survival and immortalization of cells.
In this article, we show that PAX8 activates the hTERT and hTR promoters, which in turn activate telomerase, supporting the hypothesis that PAX8 may be an important regulator of telomerase activity and of cell survival in some gliomas.Hospitals (2002)(2003)(2004)(2005).The study was approved by the institutional ethics committees and all patients provided informed consent.The inclusion and exclusion criteria have been previously published (21).Histologic diagnoses were made by consultant neuropathologists.
Telomerase activity assay.Telomerase activity was quantified using TeloTAGGG Telomerase PCR ELISA PLUS Kit (Roche Applied Science) as previously described (21).
Real-time PCR.The expression of PAX8, PAX2, hTERT, and hTR mRNA in established glioma cell lines, glioma specimens, small interfering RNA (siRNA)-treated glioma cell lines, and glioma cells overexpressing PAX8 were detected using real-time PCR.Briefly, total RNA was DNase-treated and then subjected to reverse transcription using SuperScript III and random primers (Invitrogen).For quantitative PCR, first-strand cDNA synthesized from 50 ng of RNA was used for each real-time PCR reaction.Relative quantitation of mRNA by real-time PCR using SYBR-green detection and analyzed using the Relative Expression Software Tool 2005 (REST). 4The primer sequences and PCR conditions are listed in Supplementary Table S1.
Plasmid transfection and reporter gene assays.Cells (5 Â 10 4 cells/well) were seeded into six-well plates, cultured overnight, and transfected with the hTERT or hTR promoter-luciferase plasmids (1 Ag per well of hTERT3915 or hTR2.1kb and the same copy number of hTERT255 or hTR306 with pGL2 to normalize the total amount of DNA transfected; from J. Carl Barrett; ref. 25) using FuGENE6 transfection reagent (Roche Applied Science).Cells were also transfected with either pCMV5.PAX8, pCMVPAX2b, pCMVPAX2c, or control pcDNA3 plasmids.Preparation of cell lysates and measurement of luciferase activity (Promega) have been previously described (26).Assays were performed in duplicate and repeated at least twice.
Electrophoretic mobility shift assay.Sense and antisense oligonucleotides (Supplementary Table S2) were annealed and labeled with secondgeneration DIG Oligonucleotide 3 ¶-End Labeling Kit (Roche Applied Science).After transfection with pCMV5.PAX8 for 2 days, nuclear extracts were prepared following the published protocol (27).Nuclear extract (1 Ag) was incubated with 300 ng of poly(dI-dC) in the presence or absence of unlabeled competitors for 15 min, in a 20 AL reaction containing 20 mmol/L of Tris (pH 8.0), 60 mmol/L of NaCl, 5 mmol/L of MgCl 2 , 4% glycerol, and 100 Ag of bovine serum albumin.DIG-labeled probes were then added and incubated for 15 min.The reaction mixture was then resolved on a 5% nondenaturing polyacrylamide gel in 0.5Â Tris-Glycine-EDTA buffer.DNA was transferred to a positively charged nylon membrane, Hybond-N + membrane (Amersham Pharmacia Biotech, Inc.) at 15 V overnight.The membrane was dried and then UV cross-linked for 15 min.The signal was detected using anti-Digoxigenin-AP antibody (1:10,000 dilution; Roche Applied Science) and CSPD-Star chemiluminescent substrate (Roche Applied Science).The membrane was exposed to X-ray film (Kodak) for 5 s to 15 min at room temperature.
Statistical analysis.Statistical analysis was performed using GraphPad InState (version 3.05 for Macintosh, GraphPad Software).The correlation between PAX8 mRNA expression and relative telomerase activity (RTA) was analyzed using linear regression after dividing the samples into two groups.Glioma biopsies with RTA > 350 were designated as having high telomerase activity.

Results
PAX8 expression is correlated with telomerase activity.Previously, we reported that PAX8 is expressed in a subset of gliomas (20) and that TMM is of prognostic value for patients with high-grade gliomas (21).We therefore considered the hypothesis that there is a correlation between PAX8 gene expression and TMM regulation.Results show that RTA is positively correlated with PAX8 expression level in several glioma biopsies (n = 14) that have low to moderate RTA (P < 0.0001; r 2 = 0.75; Fig. 1A).However, no correlation was observed when the RTA was very high (Fig. 1A, E).This suggests that very high RTA is also controlled by other factors.Likewise, a correlation between the PAX8 mRNA level and RTA was observed in established glioma cell lines with low to moderate telomerase activity (n = 5, r 2 = 0.97, P = 0.0025; Fig. 1B), but not in glioma cell lines with high RTA (Fig. 1B, E).These data suggest a possible role for PAX8 in transactivating one or more components of the telomerase holoenzyme (8).Glioma samples with low to moderate telomerase activity (.), glioma samples with very high telomerase activity (E).A regression of samples with low to moderate telomerase activities (dotted line ).Samples with high telomerase activities were not included in the regression.The names of the established cell lines are designated.AU, arbitrary units (log scale).The hTERT promoter contains potential PAX-binding sites.
To explore the potential transcriptional regulation of hTERT by PAX8, possible binding sites for PAX2/5/8 in the hTERT promoter were sought using the regulatory element prediction software, Consite (22).Four potential PAX2/5/8 binding sites (PB1-PB4) were found in the hTERT promoter (Fig. 2A and B), one of which is 40 bp upstream of the start codon (28).Three of the putative sites (PB1-PB3) are located in the minimal region responsible for full promoter activity (Fig. 2A; ref. 28).One of the potential PAX2/ 5/8 binding sites (PB1) overlaps with a Smad3 binding site (Fig. 2A; ref. 29), which is similar to the PAX8 binding site in the sodium/ iodide symporter gene promoter (30).The identification of these potential binding sites suggests that PAX8 could transactivate hTERT.
PAX8 binds to the hTERT promoter in vitro.To address whether PAX8 can bind these potential PAX2/5/8 binding sites in the hTERT promoter, we performed electrophoretic mobility shift assays (EMSA) using three probes spanning the four predicted binding sites (Fig. 2A).DNA-protein complexes were detected with all three hTERT probes, which could be competed out with an excess of unlabeled probe and unlabeled PAX8 consensus sequence (ConPAX8), but not by irrelevant oligonucleotides containing activator protein-1-like binding sites (PF1, see Supplementary Data; ref. 31).Probe 3 showed a weaker signal than probes 1 and 2 (Fig. 2C), as probe 3 contains only one possible binding site, whereas probes 1 and 2 contain two possible binding sites (Fig. 2C).These results were also confirmed using in vitro-translated PAX8 protein (data not shown).
Overexpression of PAX8 activates the hTERT promoter.We next tested whether PAX8 activates the hTERT promoter.pCMV5.PAX8 was cotransfected with hTERT promoter-luciferase constructs into glioma cell lines.Two hTERT luciferase reporter constructs were used.One luciferase reporter contains 255 bp of the hTERT promoter, and the other construct contains 3,915 bp of the hTERT promoter.
Because we had shown in Fig. 1 that cells with high telomerase activity might not be regulated by PAX8, we chose cell lines with high (T98G and SF268) and low (U87MG and LN18) RTA.Reporter assays were carried out 48 hours after transfection.Results showed that PAX8 expression markedly induced luciferase activity of both the short and the long hTERT promoters in a dose-dependent manner (Fig. 2D).In accordance with Fig. 1, the effect was higher in glioma cell lines with low RTA (Fig. 2D).Overexpression of PAX8 in these two cell lines activated the hTERT promoter f50-fold (Fig. 2D).In contrast, the two glioma cell lines with constitutively high RTA had lower levels of activation at f3-fold and 18-fold in SF268 and T98G, respectively (Fig. 2D).Hence, PAX8 is not only able to bind, it is also capable of activating the hTERT promoter in vitro.
Activation of the hTERT promoters by PAX8 is independent of TP53 status.It has been shown that p53 inhibits hTERT expression and telomerase activity (32,33).Also, PAX2, PAX5, and PAX8 have been reported to repress the TP53 promoter (34).To determine whether the activation of the hTERT promoter by PAX8 is partly due to its inhibitory effect on p53, we analyzed the activation of the hTERT promoters by PAX8 in the isogenic colorectal cell lines, HCT116 p53 +/+ and HCT116 p53 À/À (35).We found that PAX8 transactivated both hTERT promoters in these cell lines to a similar extent (Fig. 3).This suggests that the suppression of p53 is unlikely to play a key role in the activation of hTERT by PAX8.
The hTR promoter contains potential PAX-binding sites.As well as hTERT, hTR may also be regulated by PAX8.We therefore investigated whether PAX protein can regulate the hTR promoter (36).Analysis of the promoter, indicated that six possible PAX2/ 5/8 binding sites (PB1-PB6) are present in the short hTR promoter (À306 bp; Fig. 4A).Of these, the sequence of PB1 has the highest similarity to the consensus binding site of PAX2/5/8 and PB5 had the lowest similarity (Fig. 4B).These data suggest that PAX8 may also transactivate hTR.
PAX8 binds to the hTR promoter in vitro.To investigate whether PAX8 binds to the potential PAX2/5/8 binding sites in the hTR promoter, five probes spanning the possible binding sites in the short promoter of hTR were used for EMSA (Fig. 4C).Probes 1, 3, 4, and 5 contain one possible PAX2/5/8 binding site each, whereas probe 2 spans the region of PB2 and PB3.Specific binding was observed with probes 1, 2, and 4, as the binding signals could be competed with excess unlabeled ConPAX8 probe, but not with the irrelevant PF1 probe (Fig. 4C).By contrast, only nonspecific binding was observed with probe 3, which contains PB4, as excess unlabeled ConPAX8 did not compete out the signal (Fig. 4C).Similarly, nonspecific binding was observed with probe 5, which contains PB6, as an excess of unlabeled PF1 probe competed out  the signal (Fig. 4C).These results suggest that PAX8 could regulate the hTR promoter.
Overexpression of PAX8 activates the hTR promoter.We next tested whether PAX8 is able to activate the hTR promoter using two hTR luciferase reporter constructs.One luciferase reporter contains a 306-bp fragment of the hTR promoter, and the other construct contains a 2.1 kb hTR promoter fragment (Fig. 4D).Luciferase assays showed that PAX8 activated both of the hTR promoters in glioma cell lines (Fig. 4D).In most cases, PAX8 had a stronger activation effect on the long-hTR promoter (À2.1 kb) than on the short-hTR promoter (À306 bp), which suggests there are possible PAX8-responsive elements in the long-hTR promoter (À2.1 kb) as well (Fig. 4D).The activation of the short hTR promoter was higher in cells with low RTA at f15-fold compared with those with high RTA at f5-fold (Fig. 4D).
Effect of PAX2 isoforms on the hTERT and hTR promoters.PAX proteins of the same subgroup often recognize similar binding sites (17).It is not known whether PAX2 carries out the same function as PAX8 in glioma cells.Several alternatively spliced forms of PAX genes have been reported (23,37,38).Alternatively spliced PAX transcripts with the same DNA binding domain, but different COOH-terminals, are able to recognize identical binding sites in DNA, albeit with different transactivation capabilities (23).Hence, we also tested whether different isoforms of PAX2 act on the hTERT and the hTR promoters differently.Two of the common isoforms were chosen: PAX2b and PAX2c.PAX2b is different from PAX2c in that it does not contain exon 10 and thus has a COOH-terminal similar to PAX8 (Fig. 5A).Promoter-reporter results show (Fig. 5A  and B) that PAX2b can transactivate the hTERT promoters 3-fold to 6-fold with increasing amounts of transfected plasmids, but had no effect on hTR.In contrast, PAX2c had relatively little effect on either promoter in two different glioma cells (Fig. 5B; data not shown).Thus, it seems unlikely that PAX2 regulates telomerase.In agreement with this, PAX2 mRNA expression and the RTA levels showed no correlation in glioma biopsies and glioma cell lines (Fig. 5C).
PAX8 increases the mRNA of hTERT and hTR , as well as telomerase activity.To assess whether PAX8 also increases hTERT and hTR mRNA, we quantified the mRNA level of hTERT and hTR using real-time PCR after cells were transfected with 0.1 to 0.5 Ag of pCMV5.PAX8 for 2 days.Two-fold to 4-fold increases in hTERT and hTR mRNA expression were observed in LN18, SF268, and U87MG cells (Fig. 6A).A low activation was also observed in A172 cells, but was not included because of low transfection efficiency (data not shown).
To determine the effect of endogenous PAX8 and PAX2 on hTR and hTERT, T98G cells were transfected with either PAX8 or PAX2 siRNA and the mRNA levels of hTERT and hTR were determined by real-time PCR 24 hours posttransfection.Results show that PAX8 silencing reduces both the hTR and hTERT expression levels by two-thirds compared with the control (Fig. 6B).Whereas PAX2 silencing had no effect on hTR mRNA levels with a small but insignificant reduction in hTERT mRNA levels (P = 0.5).These results are consistent with the previous overexpression results showing that PAX8 can transactivate both hTR and hTERT promoters, yet PAX2 has a minimal effect on either promoter.Similar results were obtained in A172 and SNB75 cells (data not shown).
We next measured RTA after transfection of cells with increasing amounts of pCMV5.PAX8.Results show that PAX8 increased RTA by up to 50% (P < 0.001) with 0.1 Ag of pCMV5.PAX8 (Fig. 6C), but higher amounts of pCMV5.PAX8 reduced RTA (P < 0.001; Fig. 6C).Collectively, these data provide support for the hypothesis that PAX8 is capable of regulating telomerase in glioma cells.

Discussion
In the present study, we have found a correlation between PAX8 mRNA expression and RTA in glioma biopsies and cell lines (Fig. 1).
Both hTERT and hTR have potential PAX2/5/8 binding sites in their promoters, which were validated by EMSA (Figs. 2C and 4C).We also showed that PAX8 transactivates both the hTERT and hTR promoters (Figs.2D and 4D), increases hTERT and hTR mRNA, and increases RTA (Fig. 6A and C).Specific knockdown of PAX8 also reduces hTERT and hTR mRNA (Fig. 6B).Thus, PAX8 may be an important regulator of telomerase in some gliomas.
The correlation between PAX8 mRNA levels and RTA was only found in glioma specimens and in established glioma cell lines that had low to moderate RTA (Fig. 1).Furthermore, transactivation of the hTERT and the hTR promoters by PAX8 was much more efficient in cells with low RTA (Fig. 2D).Such data suggest that, in these gliomas, RTA is more sensitive to PAX8 levels than in others.By contrast, in gliomas with high RTA, there must be other factors controlling the activity.The logarithmic nature of the relationship between PAX8 mRNA level and RTA is consistent with other factors contributing to telomerase activity (Fig. 1A).

Cancer Research
Cancer Res 2008; 68: (14).July 15, 2008 Other factors such as transcriptional cofactors and/or chromatin might constrain the effect of PAX8 on RTA, such that very high expression of PAX8 is not able to further increase RTA and even possibly decrease RTA.Results in Figs.2D and 4D show that high expression of transfected PAX8 reduces the level of hTR and hTERT transactivation, and similarly, high levels of PAX8 decrease RTA (Fig. 6C).This could be explained by a limiting cofactor which induces squelching, such that excessive PAX8 reduces the availability of this cofactor to form complete transcription complexes.That other factors are important is evident by the fact that PAX8 failed to activate the hTERT promoter in a telomerase-negative primary cell line 5 and in two alternative lengthening of telomere-positive, PAX8-positive fibroblast cell lines (39). 5 Thus, PAX8 is not the only barrier for telomerase activation.In addition, epigenetic modifications, such as methylation, have also been shown to be critical in the control of telomerase activity (33,40,41).
Many reports have shown that PAX genes have tumor-promoting roles (19), including stimulating cell proliferation, having antiapoptotic function, and inhibiting p53 expression (20,42,43). 6At the same time, accumulating evidence suggests that telomerase has roles in addition to telomere maintenance, such as in the activation of the glycolytic pathway, cell proliferation, and inhibiting apoptosis (44)(45)(46).We have shown that PAX8 is able to activate hTERT and telomerase activity.Therefore, the antiapoptotic function of PAX genes, in addition to activation of the Bcl-2 gene (47), might also partly act through activating TERT and its antiapoptotic function.Thus, PAX8 and hTERT expression in gliomas might not only facilitate the activation of telomerase activity, but might also increase the survival function for gliomas.
Significance.The results show for the first time that the oncofetal protein, PAX8, has a novel function in telomerase activation in glioma and colorectal cancer cell lines.PAX8 coordinately regulates both hTERT and hTR.The ability of PAX genes to activate telomerase may be specific to only certain PAX genes, as it has not been observed with any other PAX genes tested thus far (48)(49)(50).The activation of hTERT expression in glioma cells by PAX8 might explain why patients with telomerase-positive gliomas have a poor prognosis (21) as both PAX8 6 and hTERT confer survival advantages for cancer cells.

Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.

Figure 1 .
Figure 1.Correlation of relative PAX8 mRNA level with RTA in glioma specimens (A) and in established glioma cell lines (B ).Glioma samples with low to moderate telomerase activity (.), glioma samples with very high telomerase activity (E).A regression of samples with low to moderate telomerase activities (dotted line ).Samples with high telomerase activities were not included in the regression.The names of the established cell lines are designated.AU, arbitrary units (log scale).

Figure 2 .
Figure 2. hTERT is a target for PAX transcription factors.A, potential PAX2/5/8 binding sites in the promoter region of the hTERT promoter.The major transcription initiation site (+1; in italics and boldface) and the translation initiation codon (ATG; highlighted in black ).The region responsible for full promoter activity (highlighted in boldface and red), potential PAX2/5/ 8 binding sites from position À255 to the translational initiation codon (underlined ), and the Smad3 binding site (highlighted in gray ).B, alignment of the PAX2, PAX5, and PAX8 potential binding sites (PB1-PB4; the first potential binding site from the distal end of the promoter is defined as PB1) in the hTERT promoter with the PAX2/5/8 consensus binding sequence and the PAX8 consensus binding sequence.Nucleotides which are the same as the PAX2/5/8 consensus sequence (highlighted in black ), and nucleotides which are the same as the PAX8 consensus sequence (highlighted in gray ).C, an EMSA was used to analyze the binding of PAX8 to digoxigenin-labeled oligonucleotide probes containing the predicted PAX binding sites found in the minimal hTERT promoter.The labeled probes were incubated with nuclear extract from U87MG cells transfected with 1 Ag of pCMV5.PAX8 for 2 d.To confirm the specificity of binding, the addition of 15-fold excess cold probe, consensus PAX8 binding sequence (ConPAX8 ), or irrelevant oligonucleotide (PF1 ) was undertaken as indicated.Comp: competitor.D, transactivation of the hTERT promoters by overexpression of PAX8 in glioma cell lines.Glioma cell lines with different telomerase activity were transfected with the hTERT luciferase constructs and varying amounts of a PAX8 expression vector.The effect of PAX8 on the short (white columns ) and the long (black columns ) promoters of hTERT was tested.Luciferase activities were measured 48 h after transfection.Relative luciferase activities were normalized to the luciferase value of the minimal promoter of hTERT in each cell line (columns, mean; bars, SE).

Figure 4 .
Figure 4. hTR is a target of the PAX transcription factor family.A, potential PAX2/5/8 binding sites in the minimal promoter region of the hTR promoter.Potential PAX2/5/8 binding sites from position À306 to the transcriptional start site (underlined ).B, alignment of the PAX2, PAX5, and PAX8 potential binding sites (PB1-PB6 ) in the hTR promoter with the PAX2/5/ 8 consensus binding sequence and the PAX8 consensus binding sequence.C, EMSA was used to analyze the binding of PAX8 to digoxigenin-labeled oligonucleotide probes containing predicted PAX binding sites found in the hTR promoter.Arrows, PAX8 binding; arrowheads, nonspecific binding.D, transactivation of the hTR promoter by overexpression of PAX8 in glioma cell lines.Relative luciferase activities were normalized to the luciferase value of the short promoter (À306) of hTR in each cell line (columns, mean; bars, SE).Left Y-axis, the transactivation level of the short hTR promoter (white columns ); right Y-axis, the transactivation level of the long hTR promoter (black columns ; refer to details in Fig. 2).

Figure 5 .
Figure 5.The effect of PAX2b or PAX2c overexpression on hTERT and hTR promoters in U87MG cell lines.A, alignment of the COOH-terminal amino acid sequences of PAX2b and PAX2c with PAX8.Highlighted regions, similarities between PAX2b and PAX8.B, the effect of PAX2b and PAX2c on hTERT and hTR promoters was analyzed by a luciferase assay.pGL2-Ctrl plasmids were transfected for comparison.Relative luciferase activities were normalized to the luciferase value of the short promoter of hTERT or hTR in each cell line (columns, mean; bars, SE).For the hTR promoters, the transactivation level of the short hTR promoter (left Y-axis ), and the transactivation level of the long hTR promoter (right Y-axis).C, correlation of relative PAX2 mRNA level with RTA in glioma tumor biopsies and established glioma cell lines.The names of the established cell lines are designated.AU, arbitrary unit (log scale).

Figure 6 .
Figure 6.PAX8 increases hTERT and hTR mRNA and telomerase activity in glioma cells.A, mRNA levels of hTERT and hTR in U87MG cells transfected with 0.1 to 0.5 Ag of pCMV5.PAX8 for 48 h.B, relative expression of hTR and hTERT in T98G cells transfected with 10 nmol/L of PAX8 or PAX2 siRNA 24 h posttransfection.mRNA levels were determined by real-time PCR and normalized to the housekeeping gene B2M (control ), expression levels are expressed relative to control transfections F SE. C, RTA of U87MG cells transfected with different amounts of PAX8 expression vector were measured.Columns, mean; bars, SE.