Most somatic human cells lack telomerase activity because they do not express the telomerase reverse transcriptase (hTERT) gene. Conversely, most cancer cells express hTERT and are telomerase positive. For most tumors it is not clear whether hTERT expression is due to their origin from telomerase positive stem cells or to reactivation of the gene during tumorigenesis. Telomerase negative cells lack detectable cytoplasmic and nuclear hTERT transcripts; in telomerase positive cells 0.2 to 6 mRNA molecules/cell can be detected. This suggests that expression is regulated by changes in the rate of hTERT gene transcription. In tumor cell lines hTERT expression behaves like a recessive trait, indicating that lack of expression in normal cells is due to one or several repressors. Studies with monochromosomal hybrids indicate that several chromosomes may code for such repressors. A number of transcription factors, tumor suppressors, cell cycle inhibitors, cell fate determining molecules, hormone receptors and viral proteins have been implicated in the control of hTERT expression; but these studies have not yet provided a clear explanation for the tumor specific expression of the hTERT gene, and the cis-acting elements which are the targets of repression in normal cells still have to be identified.
Telomerase is the enzyme required for the addition of telomeric repeats to the ends of linear chromosomes. It consists of a reverse transcriptase, TERT, that carries its own template in the form of an RNA moiety, TER. In vitro this complex can add telomeric repeats to artificial substrates. Its activity in vivo depends on other components some of which probably control the access of the enzyme to chromatid ends (Evans and Lundblad, 2000). In the absence of telomerase the telomeres of normal cells shorten by about 50 nt per cell population doubling (Counter et al., 1992; Harley et al., 1990; Huffman et al., 2000). In adult humans the enzyme is present in the germ line stem cells that give rise to mature gametes as well as in at least certain stem cell populations and in activated lymphocytes, but not in differentiated cells (Chiu et al., 1996; Wright et al., 1996). In the absence of telomerase activity human somatic epithelial cells and fibroblasts can undergo approximately 50 to 60 population doublings before telomere shortening leads to replicative senescence (see e.g. Bodnar et al., 1998). In rodents TERT expression is maintained during differentiation, and cellular senescence is not due to absence of telomerase (Russo et al., 1998). Observations on TER-deficient mice indicate that the enzyme is not required for the development and normal life span of laboratory mice in early generations (Blasco et al., 1997). However, propagation of mTER−/− mice for three or more generations leads to extensive telomere shortening and affects development and function of multiple tissues (Lee et al., 1998). In several human cell types ectopic expression of human TERT (hTERT) is sufficient to induce in vitro and in vivo telomerase activity and to ‘immortalize’ the cells, indicating that none of the other components is limiting (Bodnar et al., 1998; Morales et al., 1999; Vaziri and Benchimol, 1998; Yang et al., 1999a).
There is a striking correlation between the presence of hTERT mRNA and telomerase activity (see e.g. Ducrest et al., 2001), and this has been taken to suggest that hTERT expression is regulated through changes in the rate of transcription, but direct evidence for this is scarce. Post-transcriptional regulation of hTERT expression through alternative splicing has been observed during human development (Ulaner et al., 2001), and there have been claims that posttranslational modifications can affect TERT activity (Kang et al., 1999; Kharbanda et al., 2000; Liu et al., 2001; Yu et al., 2001), but the role of such mechanisms in tumor specific telomerase expression is, as yet, quite unclear. The finding that most tumors express hTERT and telomerase activity (Kim et al., 1994), and that in vitro transformation of telomerase negative human cells requires activation of hTERT expression (Hahn et al., 1999) indicates that maintenance of telomeres is required for the unlimited proliferative potential of tumor cells. This conclusion is supported by the finding that telomerase negative in vitro transformed cells maintain telomeres through an alternative (ALT) pathway that is based on somatic recombination (Bryan et al., 1995; Dunham et al., 2000).
For oncology the importance of understanding the mechanisms that control hTERT expression in tumors is twofold; on the one hand, it may lead to the discovery of targets for new cancer therapies, and on the other hand it might provide cis-acting regulatory elements that could contribute to tumor targeting of tumoricidal genes or viruses. Thus, it is not surprising that there have been a large number of groups that have tried to dissect the mechanisms that control hTERT expression. In this review we discuss this work, limiting ourselves to efforts to elucidate the mechanisms regulating hTERT mRNA levels, and try to explain why so far it has provided few if any conclusive answers that would be helpful to oncologists.
Maintenance of expression or activation of the hTERT gene?
Human skin or lung fibroblasts do not express hTERT, and senesce after 50 to 60 population doublings. Ectopic expression of hTERT renders these as well as endothelial cells ‘immortal’ without inducing any changes in their karyotype or other signs of transformation (Bodnar et al., 1998; Jiang et al., 1999; Morales et al., 1999; Vaziri et al., 1999; Yang et al., 1999a). There is no report of spontaneous immortalization of normal fibroblasts, but SV40 infection, by blocking the p53 and p16 dependent pathways that arrest cells when they reach senescence, extends their life span (see Duncan and Reddel, 1997 for review). These cells eventually hit a ‘crisis’ during which almost all cells die with the exception of a few transformed survivors that either maintain their telomeres by the ALT pathway (see Reddel et al., 1997 for review) or express hTERT. In this case there is no doubt that hTERT expression has been reactivated. Whether this occurs in tumors is much less clear (for a discussion of this issue see Greaves, 1996; Shay and Wright, 1996). There is evidence that some, perhaps most, tumors are derived from cells that have already gained their first alterations towards malignant transformation before undergoing differentiation, close to a stem-cell like stage when hTERT may still have been expressed. The clearest case can probably be made for colorectal carcinoma. Colorectal adenomas are derived from crypt cells some of which can express hTERT, as detected by in situ hybridization (Kolquist et al., 1998). Many adenomas themselves contain hTERT expressing cells but a proportion of them lack detectable telomerase activity (Yan et al., 2001). This may reflect the fact that most adenoma cells undergo differentiation and eventually die, while a small variable number of undifferentiated cells ensure the survival of the tumor. These may be the hTERT positive cells detected in situ.
Thus, in most carcinomas hTERT expression may not be due to reactivation of the hTERT gene but to the fact that the cells which maintain the tumor are prevented from differentiating and maintained in a stage at which their normal counterparts still express hTERT. The finding that the frequency of telomerase negative sarcomas is higher than that of carcinomas (Yan et al., 1999) suggests that sarcomas might be more frequently derived from hTERT negative cells for which there is no preferential choice of the mechanism through which they stabilize chromosome ends (Carroll et al., 1999).
Why do tumor cells need telomerase?
Telomeres are structures that prevent the ends of a linear chromosome to be mistaken for a double strand break (Godhino Ferreira and Promisel Cooper, 2001; McClintock, 1941; van Steensel et al., 1998). If these structures are disrupted, the cell attempts to repair the break and, in doing so, generates fusions between the telomeres of different chromatids. Fusions occur when the number of telomeric repeats drops below a critical level, in cells which lack telomerase (Blasco et al., 1997; Hackett et al., 2001) and do not express an ALT pathway. Thus, most tumor cells need telomerase to maintain telomeres sufficiently long to keep the incidence of chromosome fusions low. However, telomere attrition to a level at which telomeres cease to protect chromosome ends requires 50 to 60 cell doublings, and it is not clear whether the cells in a tumor have indeed undergone that many divisions, even taking into account cell loss due to differentiation and death. It seems important to consider alternative reasons for the hTERT expression by most tumors. One explanation may lie in the chromosomal instability that characterizes most cancer cells (Parshad and Sanford, 2001; see Sen, 2000 for review). At least some of this instability arises from the breakage and fusion of chromosomes. Indeed, breakage is involved in the amplification of oncogenes or genes conferring drug resistance, through breakage-fusion-bridge cycles (Coquelle et al., 1997, 1998). Although chromosome breaks can provide the cellular substrate for the selection of more aggressive tumor cells, they will also give rise to non-viable cells. One way to keep these processes in check is through de novo addition, by telomerase, of telomeres to the ends of broken chromosomes (Friebe et al., 2001; Hande et al., 1998; Varley et al., 2000). This would mean that premalignant cells, which express hTERT, have an advantage over the others not only when cells have undergone more than 50 to 60 divisions, but at a much earlier stage when chromosome breakage becomes frequent.
Is hTERT expression regulated by changes in the level of gene transcription?
As pointed out above there is a very strong correlation between telomerase expression and the presence of detectable hTERT mRNA (Meyerson et al., 1997; Nakamura et al., 1997). We have compared the numbers of hTERT molecules per cell, determined by quantitative RT–PCR, in a number of cell lines from different tissue origins (Ducrest et al., 2001). In all telomerase positive cells hTERT transcripts are detectable but rare (0.2 to 6/cell) whereas no transcripts (<0.004/cell) could be detected in telomerase negative cells. This correlation has been widely assumed to reflect regulation of hTERT expression via control of the rate of transcription. But it is equally compatible with regulation of transcript processing or changes in the mRNA half-life. Although a considerable number of transcription factors have been implicated in the control of hTERT expression, direct evidence that hTERT gene transcription is regulated is scarce. Specifically it is unclear whether the tumor specific expression of hTERT is controlled at the level of transcription. The finding that activation of a c-Myc-estrogen receptor ligand binding domain fusion can increase hTERT mRNA levels in the absence of protein synthesis shows that ectopic c-Myc can indeed directly stimulate transcription of the gene (Greenberg et al., 1999; Oh et al., 2000; Wu et al., 1999). We will discuss the biological role of c-Myc in hTERT regulation below.
Run-on nuclear experiments which measure the average loading of RNA-polymerase molecules on the gene, have been reported for one leukemia cell line (U937) (Gunes et al., 2000). This study indicated that in these cells hTERT is regulated at the level of transcription rather than RNA stability. We have made attempts to obtain similar evidence for a tumor cell line derived from fibrosarcoma (HT1080) that contains relatively high numbers of hTERT transcripts among the cell lines screened by us, and have been unable to detect run-on transcription signals above background. The probable reason for this failure is that the rate of transcription is too low in HT1080 cells to be detectable by this approach. Comparing the levels of spliced cytoplasmic mRNA with that of intron-containing nuclear transcripts in different telomerase positive and negative cell lines, we observed that telomerase negative cells did not contain detectable levels (<0.004 molecules/cell) of either cytoplasmic mRNA or nuclear transcripts, whereas telomerase positive cells contained both transcript forms (Ducrest et al., 2001). These results clearly suggest that hTERT mRNA levels are indeed controlled at the level of gene transcription, but they do not exclude that regulation involves changes in the efficiency of nuclear processing of primary transcripts.
Possible models of hTERT regulation
In a sense hTERT behaves like a protooncogene; abnormal maintenance or reactivation of expression contributes to tumorigenesis. Thus, one would expect that genomic changes that can lead to improper expression of protooncogenes, such as translocations that include the regulatory regions, would also be found in the hTERT genes of tumors.
Indeed, there is one study (Horikawa and Barrett, 2001) suggesting that the integration of the hepatitis B viral genome into the 5′ flanking region of the hTERT gene might induce its expression in a hepatocellular carcinoma. We have found no evidence for rearrangements in the 5′ flanking region and the 5′ half of the gene (−10 to +25 kb) screening a number of cell lines of divers origin. The second intron of the hTERT gene contains a meiotically unstable minisatellite with several putative binding sites for c-Myc (Szutorisz et al., 2001; Wu et al., 1999). Size rearrangements of that minisatellite are not required for telomerase expression in colon carcinomas (Szutorisz et al., 2001). In 31 of 33 colon carcinomas that were heterozygous for the polymorphic minisatellite the 1 : 1 ratio of hTERT alleles was maintained, indicating that there had been no gene amplification in these tumors. In the two remaining tumors there was a change compared to normal tissue from the same patient, compatible with amplification of one hTERT allele. Amplification of the hTERT gene was also detected in another study, in 20% of primary tumors and 40% human cancer derived cell lines (Zhang et al., 2000). Amplification may be the result of selection for higher expression of an active hTERT gene. It might also lead to the expression of an inactive gene as a consequence of the genomic rearrangements that give rise to amplification, or through titration of a gene specific repressor.
Another modification that might affect hTERT expression is DNA methylation. Turning off the expression of tumor suppressor genes or genes involved in DNA repair, through methylation of their promoter, can contribute to carcinogenesis. Comparison of the methylation status of the hTERT promoter in telomerase positive and negative cells has not provided any compelling clues that this type of modification controls tumor specific hTERT expression (Dessain et al., 2000; Devereux et al., 1999).
hTERT expression due to cis-acting gene rearrangements should behave like a dominant trait. Dominant expression would also be likely if demethylation of the hTERT promoter were the mechanism through which hTERT expression is activated in tumors. However, so far no cross in which hTERT expression is dominant has been reported. On the other hand there is a number of tumor lines in which hTERT expression behaves like a recessive trait; expression is extinguished in hybrids with telomerase negative cells or by transfer of a single chromosome from a normal cell (Table 1) (Bryan et al., 1995; Cuthbert et al., 1999; Horikawa et al., 1998; Nishimoto et al., 2001). This suggests that hTERT expression in normal cells is repressed by a mechanism which is no longer functional in tumors. The simplest model that accounts for these observations is that hTERT transcription is under the control of a repressor, absent in cancer cells, that acts via a cis-acting element in the hTERT gene. Note, that the repressor may not itself be a sequence specific DNA-binding protein, but could be a co-repressor interacting with a transcription factor. The data are equally compatible with a model according to which the repressor controls a gene coding for an obligatory activator of hTERT transcription, and so on. The finding that single normal chromosomes can repress hTERT expression in tumors has led to attempts to clone the genes coding for such repressors, by positional cloning.
How many hTERT repressors are there?
Table 1 lists the chromosome transfer experiments that have addressed the question of hTERT regulation. The data summarized are not homogeneous, and different studies testing the same chromosome have not always used the same chromosome donor cells. It should also be kept in mind that a normal chromosome may undergo changes in the donor cells. This might explain that chromosome 6 represses hTERT in the cervical carcinoma line SiHa in one study but fails to do so in another. Alternatively, the different result may reflect changes in the cell line. Given these limitations the studies listed in Table 1 strongly suggest that there is no single chromosome that represses hTERT expression in all cells. Chromosome 3, e.g. represses hTERT expression in several but not all of the recipient lines tested. This is not unexpected; even if there were a single molecular complex that is responsible for the repression of the hTERT gene in normal cells, mutations in both copies of any gene coding for a component of the complex should lead to inactivation of the repressor and expression of hTERT. It would certainly be interesting to determine whether a large scale chromosome screen would reveal patterns, e.g. consistent repression of hTERT expression by chromosome 6 in HPV16-transformed tumors. The available data do not permit to detect such patterns. Transfer of chromosomes from irradiated donor cells can be used for attempts to positionally clone a putative hTERT repressor gene. The chromosome for which this approach is most advanced is chromosome 3. Upon introduction of a normal chromosome 3, two renal, one breast, and one cervical carcinoma line ceased to express hTERT. Two groups using either a kidney renal carcinoma (Tanaka et al., 1998) or a breast cancer derived line (Cuthbert et al., 1999) as recipients have narrowed the region that confers repression to p14.2–21.1. This region overlaps with a segment of chromosome 3 that undergoes frequent LOH in breast cancer (Maitra et al., 2001). LOH and deletions of smaller parts of 3p have been identified in breast, cervix, colon, lung, and renal carcinomas (Kok et al., 1997).
In a single study both chromosome 3 and 4 have been found to shut off hTERT expression in HeLa cells. This suggests multiple independent pathways of repression. Since mutations affecting a repressive pathway are recessive, activation of hTERT expression through such mutations would be expected to be an extremely rare event. It might explain why spontaneous immortalization of normal fibroblasts has never been observed.
Other approaches to study regulation of hTERT expression
Screening candidate molecules
The mapping and cloning of genes on normal chromosomes that shut off hTERT expression in tumor cells is one approach towards the elucidation of the regulation of hTERT expression. Other, complementary approaches consist in (1) the testing of candidate molecules for their effect on the expression of the endogenous hTERT gene, or (2) attempts to identify the cis-acting elements in the hTERT gene that control its expression. The former approach is based on guesses as to what molecules might be involved in hTERT regulation which can be tested either through the ectopic expression of such putative positive regulators in hTERT negative cells, or through the expression of dominant negative version of such molecules in hTERT expressing cells. The latter is a better approach that can provide informative data even if the results are negative. Candidates include molecules whose abnormal expression in tumor cells prevents their differentiation, such as c-Myc, TCF or Notch. The second approach aims at the identification of cis-acting regulatory sequences in or near the hTERT gene through experiments using reporter gene constructs and including in vitro assays for DNA binding proteins, nuclease hypersensitivity, in vivo footprinting assays and ChromatinIP.
Numerous molecules, including transcription factors, regulators of differentiation and the cell cycle, and proteins of viruses implicated in tumorigenesis, have been proposed to regulate hTERT expression. We have attempted to summarize the most relevant findings in Table 2 and Figure 1, without being exhaustive in our literature citations. Many studies were based on the ectopic expression of positive regulators. The interpretation of such experiments is often difficult. An example is provided by the studies on the effect of c-Myc on hTERT expression. The published data show that overexpression of c-Myc can increase the level of hTERT mRNA in B-cell lines or induces its appearance in fibroblasts. This effect does not depend on protein synthesis and is therefore likely to be due to a direct action of c-Myc protein on the hTERT gene (Greenberg et al., 1999; Oh et al., 2000; Wu et al., 1999). Mad, the antagonist of c-Myc was shown to be a potential repressor of hTERT. Mad was a candidate repressor identified in a gene screen for hTERT regulators (Oh et al., 2000) and a rise in endogenous Mad RNA and protein levels was inversely correlated with hTERT RNA levels (Gunes et al., 2000; Oh et al., 2000; Xu et al., 2001). Finally, while c-Myc protein was found associated with the hTERT gene in vivo in telomerase-positive promyelocytic leukemia HL60 cells as determined in chromatin immunoprecipitation assays (Xu et al., 2001), differentiation of these cells by DMSO led to downregulation of hTERT, loss of association with c-Myc and binding of the c-Myc antagonist Mad1. These results show that the c-Myc/Mad regulatory network can regulate hTERT expression, but the role of this network in tumor specific hTERT expression is not yet clear. Deregulation of the c-Myc/Mad balance is unlikely to be sufficient for the activation of the hTERT gene in cancers, for several reasons: (1) In most cases overexpression of c-Myc is expected to behave like a dominant trait in somatic cell crosses, unlike what has been observed for hTERT expression. (2) In exponentially growing fibroblasts c-Myc is expressed at lower levels than in tumor derived cell lines (Gewin and Galloway, 2001; Kyo et al., 2000; Oh et al., 2000), and declines even further when fibroblasts are serum deprived. Restimulation with serum induces a transient, high level of c-Myc and downregulation of Mad (Grandori et al., 2000; Obaya et al., 1999), but there is no evidence that this change is sufficient to induce hTERT expression. (3) Overexpression of HPV16 E7, which is important for immortalization of keratinocytes, induces high level of c-Myc protein but is unable to activate telomerase expression (Gewin and Galloway, 2001; Veldman et al., 2001). (4) In the breast cancer derived cell line 21NT chromosome 3 transfer leads to immediate repression of the hTERT gene but expression of c-Myc, Mad1 and c-Myc target genes remained unchanged (Ducrest et al., 2001). Therefore, the putative repressor on chromosome 3 does not regulate hTERT through c-Myc or one of its coregulators. In conclusion, it seems likely that normal changes in the c-Myc/Mad ratio control hTERT transcription in cells in which the gene is not ‘closed’ by one or several repressors, but that the levels of c-Myc in most tumors are not high enough to overcome repression. One obvious possibility is that in normal cells competent to express the gene c-Myc links hTERT expression to the proliferative status of the cell. Other genes involved in the control of cell cycle progression have been suggested to repress hTERT expression such as p53, p16, p21, and E2F-1 (Table 2 and Figure 1). However, the effect of these genes on hTERT expression remains ambiguous. The best case can be made for p53, which was shown to downregulate hTERT expression. This effect seems to be independent of p53 induced cell cycle arrest and apoptosis (Kanaya et al., 2000; Kusumoto et al., 1999; Xu et al., 2000). Another case in which hTERT can be regulated independently of differentiation and/or growth inhibition is the acute promyelocytic leukemia cell line NB4-R1, in which treatment with retinoic acid dowregulates hTERT without inducing maturation (Pendino et al., 2001).
As suggested above hTERT expression in most carcinomas may not be due to a reactivation of the hTERT gene but reflect the advantage, during tumor progression of cells in which differentiation is partially or completely blocked and, as a consequence, hTERT expression maintained. This view would predict that pathways which control cell differentiation and which are frequently deregulated in cancer, such as the Notch and the Wnt pathways may be implicated in hTERT regulation. We have found that in the breast cancer cell line, 21NT, overexpression of the intracellular part of the Notch 1 protein increases the expression levels of hTERT transcripts as well as of HES-1, a known Notch 1 target (A Ducrest; unpublished data). Similarly, arguing that TCF activity may be required for hTERT expression in colon carcinoma cells we have determined the levels of hTERT mRNA in four colon carcinoma cell lines carrying tetracycline inducible constructs coding for dominant negative version of TCF1 or TCF4. These lines were prepared by Marc van de Wetering in the laboratory of Hans Clevers. Tetracycline treatment of such cells leads to a significant down-regulation of a number of TCF target genes expressed in colon carcinomas, but had no effect on hTERT transcript levels which were comparable to that in control cells from the same tumors lacking the dominant negative TCF constructs. These results quite strongly argue that TCF does not play a role, direct or indirect, in controlling hTERT expression in colon carcinomas.
In estrogen-targeted tissues, such as endometrium (Kyo et al., 1997; Saito et al., 1997; Takakura et al., 1999), prostate (Meeker et al., 1996) and epithelial cells with high renewal potential (Bednarek et al., 1998) estrogen-responsive cells may be more prone to form tumors (Hilakivi-Clarke, 2000; Liehr, 2000) because they are telomerase positive. Estrogen was shown to activate hTERT promoter constructs through estrogen responsive elements (ERE) in the hTERT 5′ flanking region. This activation was dependent on the presence of estrogen receptor-α. Genomic footprinting indicated that one ERE element, 950 bp upstream of the translation start site, is occupied in vivo in cells expressing, but not in cells lacking, the estrogen-receptor-α (Misiti et al., 2000). This is in agreement with the finding that tamoxifen, an antagonist of estrogen, reduces telomerase activity in the breast cancer cell line MCF-7 cells (Aldous et al., 1999). Since in this line estrogen also increases c-Myc levels (Kyo et al., 1999), c-Myc may contribute to activation of hTERT transcription.
On the use of hTERT-reporter constructs
There have been many attempts to identify cis-acting regulatory elements in the hTERT gene through the use of reporter constructs (see Table 2 and Figure 1). The main a priori limitation of this approach is that it makes assumptions on the location of the regulatory elements, which can be at considerable distance, 3′ or 5′ of the gene, or in introns. Furthermore, certain regulatory elements might not work outside of their endogenous context. Therefore the use of the basic reporter construct should be validated in experiments testing whether it contains the cis-acting elements controlling the expression of the endogenous gene, e.g. by transfection into appropriate cell lines. Claims that hTERT reporter expression reflects that of the endogenous gene have been based on the comparison of reporter expression in normal cells with that in various cell lines. However, in vitro transformed or tumor cells differ from normal cells in many respects that can affect the rate of gene transcription in ways which are unrelated to gene specific regulation. To solve this problem we (Ducrest et al., 2001) have compared the expression of a series of reporter constructs containing the hTERT promoter and up to 7.5 kb of 5′ flanking region in two SV40 transformed fibroblast lines. One of these is telomerase positive, whereas the other uses the ALT pathway and contains no detectable hTERT transcripts. All reporter constructs were more strongly expressed in either line than in normal fibroblasts, and there were no significant differences between the activity of any of the reporters in the telomerase positive and the ALT line. Knight et al. (2001) have also reported hTERT promoter activity in an ALT cell line SUSM-1 when using another reporter containing 1.7 kb of the hTERT flanking region. Even more strikingly, we observed no differences in the expression of the same hTERT reporters when we compared them in a breast carcinoma line and its derivatives in which transfer of a single normal chromosome 3 has reduced hTERT mRNA by at least 30-fold, to undetectable levels (Ducrest et al., 2001). Thus, by these stringent criteria the validation of hTERT reporters containing the longest 5′ flanking segment tested so far has completely failed, and the significance of the results obtained with similar constructs in other cells (see Table 2) has to be assessed in the light of this failure.
Of course, this does not mean that regulatory sites identified in constructs expression which does not mimic that of the endogenous gene have no role in the regulation of the latter, but without strong additional evidence such identifications provide only very weak arguments. The finding that an element identified in this way indeed binds a transcription factor that might be implicated in the regulation in vitro adds very little weight to the argument. Strong evidence that binding of a transcription factor to a putative regulatory site plays a role in the control of gene expression requires demonstration that the factor occupies the site in vivo, most convincingly by ChromatinIP with antibodies against the putative regulator. But even such experiments cannot, by themselves, prove that the transcription factor in question controls the difference in hTERT gene expression in normal versus tumor cells. It is quite possible that certain transcription factor binding sites are indeed occupied in hTERT expressing but not in telomerase negative cells, and that occupation is required for hTERT transcription. But occupancy may reflect that fact that in cells competent to express the gene these sites are ‘open’ i.e. accessible to the transcription factor, due to chromatin alterations that depend on other proteins which bind elsewhere and are higher up in the hierarchy of control.
Screening for changes in hTERT chromatin
The search for differences between the conformation of the chromatin containing the hTERT gene in hTERT expressing and non-expressing cells provides a complementary approach to the identification of cis-acting elements. The classical method used is to screen the locus of interest for sites with differential sensitivity to nucleases such as DNaseI or Micrococcal nuclease (MNase). In numerous instances the activity of a regulatory element correlates with the presence of a nuclease hypersensitive site or region at or near the element. Compared to ChromatinIP, this type of analysis has the advantage that it can be applied to very large genomic segments without previous assumption about the possible localization of regulatory elements, but it has the disadvantage that there are no strict rules describing the relationship between, say, transcription factor occupancy of a regulatory site and its nuclease sensitivity. Thus, lack of nuclease sensitive sites in a segment does not exclude that it plays a regulatory role. Application of the technique to the hTERT gene has to face another uncertainty; as discussed above the rate of hTERT gene transcription is probably very low (Ducrest et al., 2001), and even in a cloned hTERT positive cell line not all cells may transcribe the gene at a given moment. This may – or may not – mean that important regulatory elements in the gene are not always occupied, and that the corresponding nuclease hypersensitive sites are invisible in the background of chromatin from non-transcribed genes. Nevertheless, comparison of different telomerase positive and negative cell lines points to the existence of two nuclease sensitive sites in the second intron of telomerase expressing cells, and the significance of these sites has been validated by the stringent type of criteria outlined above for reporter construct analysis (H Szutorisz, manuscript in preparation). It remains to be seen whether these sites are the primary targets of the molecules that induce hTERT transcription in tumors, or whether these chromatin alterations are the downstream consequence of the activity of cis-acting elements elsewhere in the gene.
One impression that emerges from this review is that in spite of considerable efforts by many groups, our understanding of the mechanisms that are responsible for the tumor specific expression of the human TERT gene is still very poor. This raises two questions. On the one hand one has to ask what new or at least modified approaches are most likely to be more successful than the attempts carried out so far, and on the other one is lead to consider the possibility that the models which determine the choice of methods are inappropriate or wrong.
At this time it seems that the approach which is most likely to provide insight into the regulation of hTERT expression is the positional cloning of genes on chromosomes that shut off hTERT expression upon microcell mediated transfer into tumor cell lines. However, it is by no means certain that such genes once they have been identified provide immediate clues as to the mechanisms through which they affect hTERT expression, and to unravel these mechanisms it would certainly be extremely useful if not essential to have a reporter system which does mimic the expression pattern of the endogenous hTERT gene according to the stringent criteria outlined above. To build such a system may require the use of much larger genomic segments as they are available, e.g., in BAC clones. BAC clones containing the hTERT gene are available but their sequence is not yet publicly accessible. Reporter constructs based on BAC clones of other genes have been successfully used for the study of regulation, but the technical investment required is not trivial, and one needs to take into account the risk that the experiments fail because of the very low level of hTERT transcription.
If a reporter system that faithfully reproduces the tumor-specific regulation of the hTERT gene were available, it might be informative to determine its expression pattern in transgenic mice. As pointed out earlier, the TERT of the mouse (mTERT) and other rodents is not shut off in differentiated somatic cells (Russo et al., 1998). This difference between rodents and man may reflect changes in the cis-acting elements or in the expression of transacting factors. If hTERT gene expression in the mouse resembles that in man this would argue strongly that repression of hTERT expression during differentiation is due to differences in cis-acting elements only. What evolutionary pressure may have led to the somatic repression of hTERT expression? A simple idea is that this may be related to the species' life-span; in species that reach reproductive age late, repression of telomerase activity which provides an important barrier to malignant disease should confer a stronger selective advantage than in species with a short life-span. Not enough species have been analysed to allow evaluation of this hypothesis.
There are several aspects of hTERT expression that are puzzling and apparently contradictory. The finding that immortalization of normal fibroblasts by spontaneous activation of hTERT expression has never been observed, and that it is a rare event even after viral transformation, is hard to reconcile with the finding that it is quite easy to turn on hTERT expression in normal cells, through overexpression of c-Myc or treatment with an inhibitor of histone deacetylases (Cong and Bacchetti, 2000; Takakura et al., 2001; Xu et al., 2001). The indication, from monochromosomal tumor cell hybrids, that there are different genetic loci which can shut down hTERT expression, suggests that perhaps repression of hTERT is due to diffuse mechanisms that affect the chromatin structure in and around the hTERT gene, rather than to a few well defined target sites of sequence specific repressors or activators. In this context it may be relevant that the hTERT gene is close to the telomere of the short arm of chromosome 5. This raises the possibility that the gene is subject to telomeric repression which has recently been shown to exist in human cells (Baur et al., 2001). The precise position of the hTERT gene has not yet been determined (Bryce et al., 2000). It will be interesting to test whether expression of other genes close to the telomere of chromosome 5p correlates with that of hTERT.
Aldous WK, Marean AJ, DeHart MJ, Matej LA, Moore KH . 1999 Cancer 85: 1523–1529
Artandi SE, Chang S, Lee SL, Alson S, Gottlieb GJ, Chin L, DePinho RA . 2000 Nature 406: 641–645
Backsch C, Wagenbach N, Nonn M, Leistritz S, Stanbridge E, Schneider A, Durst M . 2001 Genes Chromo. Cancer 31: 196–198
Baur JA, Zou Y, Shay JW, Wright WE . 2001 Science 292: 2075–2077
Bednarek AK, Chu YL, Aldaz CM . 1998 Oncogene 16: 381–385
Blasco MA, Lee HW, Hande MP, Samper E, Lansdorp PM, DePinho RA, Greider CW . 1997 Cell 91: 25–34
Bodnar AG, Ouellette M, Frolkis M, Holt SE, Chiu C-P, Morin GB, Harley CB, Shay JW, Lichtsteinter S, Wright WE . 1998 Science 279: 349–352
Braunstein I, Cohen-Barak O, Shachaf C, Ravel Y, Yalon-Hacohen M, Mills GB, Tzukerman M, Skorecki KL . 2001 Cancer Res. 61: 5529–5536
Bryan TM, Englezou A, Gupta J, Bacchetti S, Reddel RR . 1995 EMBO J. 14: 4240–4248
Bryce LA, Morrison N, Hoare SF, Muir S, Keith WN . 2000 Neoplasia 2: 197–201
Carroll T, Maltby E, Brock I, Royds J, Timperley W, Jellinek D . 1999 J. Pathol. 188: 395–399
Chiu CP, Dragowska W, Kim NW, Vaziri H, Yui J, Thomas TE, Harley CB, Lansdorp PM . 1996 Stem Cells 14: 239–248
Cong YS, Bacchetti S . 2000 J. Biol. Chem. 275: 35665–35668
Coquelle A, Pipiras E, Toledo F, Buttin G, Debatisse M . 1997 Cell 89: 215–225
Coquelle A, Toledo F, Stern S, Bieth A, Debatisse M . 1998 Mol. Cell. 2: 259–265
Counter CM, Avilion AA, LeFeuvre CE, Stewart NG, Greider CW, Harley CB, Bacchetti S . 1992 EMBO J. 11: 1921–1929
Crowe DL, Nguyen DC, Tsang KJ, Kyo S . 2001 Nucleic Acids Res. 29: 2789–2794
Cuthbert AP, Bond J, Trott DA, Gill S, Broni J, Marriott A, Khoudoli G, Parkinson EK, Cooper CS, Newbold RF . 1999 J. Natl. Cancer Inst. 91: 37–45
Cuthbert AP, Trott DA, Ekong RM, Jezzard S, England NL, Themis M, Todd CM, Newbold RF . 1995 Cytogene. Cell Genetics 71: 68–76
Dessain SK, Yu H, Reddel RR, Beijersbergen RL, Weinberg RA . 2000 Cancer Res. 60: 537–541
Devereux TR, Horikawa I, Anna CH, Annab LA, Afshari CA, Barrett JC . 1999 Cancer Res. 59: 6087–6090
Ducrest AL, Amacker M, Mathieu YD, Cuthbert AP, Trott DA, Newbold RF, Nabholz M, Lingner J . 2001 Cancer Res. 61: 7594–7602
Duncan EL, Reddel RR . 1997 Biochemistry-Moscow 62: 1263–1274
Dunham MA, Neumann AA, Fasching CL, Reddel RR . 2000 Nat. Genet. 26: 447–450
Evans SK, Lundblad V . 2000 J. Cell Science 113: 3357–3364
Friebe B, Kynast RG, Zhang P, Qi L, Dhar M, Gill BS . 2001 Chromosome Res 9: 137–146
Fujimoto K, Kyo S, Takakura M, Kanaya T, Kitagawa Y, Itoh H, Takahashi M, Inoue M . 2000 Nucleic Acids Res. 28: 2557–2562
Gewin L, Galloway DA . 2001 J. Virol. 75: 7198–7201
Godhino Ferreira M, Promisel Cooper J . 2001 Mol. Cell. 7: 55–63
Grandori C, Cowley SM, James LP, Eisenman R.N . 2000 Annu. Rev. Cell. Dev. Biol. 16: 653–699
Greaves M . 1996 Trends Genet. 12: 127–128
Greenberg RA, O'Hagan RC, Deng H, Xiao Q, Hann SR, Adams RR, Lichtsteiner S, Chin L, Morin GB, DePinho RA . 1999 Oncogene 18: 1219–1226
Gunes C, Lichtsteiner S, Vasserot AP, Englert C . 2000 Cancer Res. 60: 2116–2121
Hackett JA, Feldser DM, Greider CW . 2001 Cell 106: 275–286
Hahn WC, Stewart SA, Brooks MW, York SG, Eaton E, Kurachi A, Beijersbergen RL, Knoll JH, Meyerson M, Weinberg RA . 1999 Nat. Med. 5: 1164–1170
Hande MP, Lansdorp PM, Natarajan AT . 1998 Mutat. Res. 404: 205–214
Harley CB, Futcher AB, Greider CW . 1990 Nature 345: 458–460
Hensler PJ, Annab LA, Barrett JC, Pereira-Smith OM . 1994 Mol. Cell. Biol. 14: 2291–2297
Hilakivi-Clarke L . 2000 Cancer Res. 60: 4993–5001
Horikawa I, Barrett JC . 2001 J. Natl. Cancer Inst. 93: 1171–1173
Horikawa I, Cable PL, Afshari C, Barrett JC . 1999 Cancer Res. 59: 826–830
Horikawa I, Oshimura M, Barrett JC . 1998 Mol. Carcinog. 22: 65–72
Huffman KE, Levene SD, Tesmer VM, Shay JW, Wright WE . 2000 J. Biol. Chem. 275: 19719–19722
Jiang XR, Jimenez G, Chang E, Frolkis M, Kusler B, Sage M, Beeche M, Bodnar AG, Wahl GM, Tlsty TD, Chiu CP . 1999 Nat. Genet. 21: 111–114
Kanaya T, Kyo S, Hamada K, Takakura M, Kitagawa Y, Harada H, Inoue M . 2000 Clin. Cancer Res. 6: 1239–1247
Kang SS, Kwon T, Kwon DY, Do SI . 1999 J. Biol. Chem. 274: 13085–13090
Kharbanda S, Kumar V, Dhar S, Pandey P, Chen C, Majumder P, Yuan ZM, Whang Y, Strauss W, Pandita TK, Weaver D, Kufe D . 2000 Curr. Biol. 10: 568–575
Kim NW, Piatyszek MA, Prowse KR, Harley CB, West MD, Ho PL, Coviello GM, Wright WE, Weinrich SL, Shay JW . 1994 Science 266: 2011–2015
Kitagawa Y, Kyo S, Takakura M, Kanaya T, Koshida K, Namiki M, Inoue M . 2000 Clin. Cancer Res. 6: 2868–2875
Knight JS, Cotter MA, Robertson ES . 2001 J. Biol. Chem. 276: 22971–22978
Koi M, Shimizu M, Morita H, Yamada H, Oshimura M . 1989 Jpn. J. Cancer Res. 80: 413–418
Kok K, Naylor SL, Buys CH . 1997 Adv. Cancer Res. 71: 27–92
Kolquist KA, Ellisen LW, Counter CM, Meyerson M, Tan LK, Weinberg RA, Haber DA, Gerald WL . 1998 Nat. Genet. 19: 182–186
Kugoh H, Mitsuya K, Meguro M, Shigenami K, Schulz TC, Oshimura M . 1999 DNA Res. 6: 165–172
Kusumoto M, Ogawa T, Mizumoto K, Ueno H, Niiyama H, Sato N, Nakamura M, Tanaka M . 1999 Clin. Cancer Res. 5: 2140–2147
Kyo S, Takakura M, Kanaya T, Zhuo W, Fujimoto K, Nishio Y, Orimo A, Inoue M . 1999 Cancer Res. 59: 5917–5921
Kyo S, Takakura M, Kohama T, Inoue M . 1997 Cancer Res. 57: 610–614
Kyo S, Takakura M, Taira T, Kanaya T, Itoh H, Yutsudo M, Ariga H, Inoue M . 2000 Nucleic Acids Res. 28: 669–677
Lee HW, Blasco MA, Gottlieb GJ, Horner JW, Greider CW, Depinho RA . 1998 Nature 392: 569–574
Liehr JG . 2000 Endocr. Rev. 21: 40–54
Liu K, Hodes RJ, Weng N . 2001 J. Immunol. 166: 4826–4830
Maitra A, Wistuba II, Washington C, Virmani AK, Ashfaq R, Milchgrub S, Gazdar AF, Minna JD . 2001 Am. J. Pathol. 159: 119–130
McClintock B . 1941 Genetics 26: 234–282
Meeker AK, Sommerfeld HJ, Coffey DS . 1996 Endocrinology 137: 5743–5746
Meyerson M, Counter CM, Eaton EN, Ellisen LW, Steiner P, Dickinson Caddle S, Ziaugra L, Beijershergen RL, Davidoff MJ, Liu Q, Bacchetti S, Haber DA, Weinberg RA . 1997 Cell 90: 785–795
Misiti S, Nanni S, Fontemaggi G, Cong YS, Wen J, Hirte HW, Piaggio G, Sacchi A, Pontecorvi A, Bacchetti S, Farsetti A . 2000 Mol. Cell. Biol. 20: 3764–3771
Morales CP, Holt SE, Ouellette M, Kaur KJ, Yan Y, Wilson KS, White MA, Wright WE, Shay JW . 1999 Nat. Genet. 21: 115–118
Nakabayashi K, Ogino H, Michishita E, Satoh N, Ayusawa D . 1999 Exp. Cell Res. 252: 376–382
Nakamura TM, Morin GB, Chapman KB, Weinrich SL, Andrews WH, Lingner J, Harley CB, Cech TR . 1997 Science 277: 955–959
Ning Y, Lovell M, Taylor L, Pereira-Smith OM . 1992 Cytogenet. Cell Genet. 60: 79–80
Nishimoto A, Miura N, Horikawa I, Kugoh H, Murakami Y, Hirohashi S, Kawasaki H, Gazdar AF, Shay JW, Barrett JC, Oshimura M . 2001 Oncogene 20: 828–835
Obaya AJ, Mateyak MK, Sedivy JM . 1999 Oncogene 18: 2934–2941
Oh S, Song Y, Yim J, Kim TK . 1999a J. Biol. Chem. 274: 37473–37478
Oh S, Song YH, Kim UJ, Yim J, Kim TK . 1999b Biochem. Biophys. Res. Commun. 263: 361–365
Oh S, Song YH, Yim J, Kim TK . 2000 Oncogene 19: 1485–1490
Oh ST, Kyo S, Laimins LA . 2001 J. Virol. 75: 5559–5566
Ohmura H, Tahara H, Suzuki M, Ide T, Shimizu M, Yoshida MA, Tahara E, Shay JW, Barrett JC, Oshimura M . 1995 Jpn. J. Cancer Res. 86: 899–904
Oshimura M, Barrett JC . 1997 Eur. J. Cancer 33: 710–715
Parshad R, Sanford KK . 2001 Crit. Rev. Oncol. Hematol. 37: 87–96
Pendino F, Flexor M, Delhommeau F, Buet D, Lanotte M, Segal-Bendirdjian E . 2001 Proc. Natl. Acad. Sci. USA 98: 6662–6667
Reddel RR, Bryan TM, Murnane JP . 1997 Biochemistry 62: 1254–1262
Russo I, Silver AR, Cuthbert AP, Griffin DK, Trott DA, Newbold RF . 1998 Oncogene 17: 3417–3426
Saito T, Schneider A, Martel N, Mizumoto H, Bulgay-Moerschel M, Kudo R, Nakazawa H . 1997 Biochem. Biophys. Res. Commun. 231: 610–614
Sen S . 2000 Curr. Opin. Oncol. 12: 82–88
Shay JW, Wright WE . 1996 Trends Genet. 12: 129–131
Steenbergen RD, Kramer D, Meijer CJ, Walboomers JM, Trott DA, Cuthbert AP, Newbold RF, Overkamp WJ, Zdzienicka MZ, Snijders PJ . 2001 J. Natl. Cancer Inst. 93: 865–872
Szutorisz H, Palmqvist R, Roos G, Stenling R, Schorderet DF, Reddel R, Lingner J, Nabholz M . 2001 Oncogene 20: 2600–2605
Takakura M, Kyo S, Kanaya T, Hirano H, Takeda J, Yutsudo M, Inoue M . 1999 Cancer Res. 59: 551–557
Takakura M, Kyo S, Sowa Y, Wang Z, Yatabe N, Maida Y, Tanaka M, Inoue M . 2001 Nucleic Acids Res. 29: 3006–3011
Tanaka H, Horikawa I, Kugoh H, Shimizu M, Barrett JC, Oshimura M . 1999 Mol. Carcinog. 25: 249–255
Tanaka H, Shimizu M, Horikawa I, Kugoh H, Yokota J, Barrett JC, Oshimura M . 1998 Genes Chrom. Cancer 23: 123–133
Ulaner GA, Hu JF, Vu TH, Giudice LC, Hoffman AR . 2001 Int. J. Cancer 91: 644–649
van Steensel B, Smogorzewska A, de Lange T . 1998 Cell 92: 401–413
Varley H, Di S, Scherer SW, Royle NJ . 2000 Am. J. Hum. Genet. 67: 610–622
Vaziri H, Benchimol S . 1998 Curr. Biol. 8: 279–282
Vaziri H, Squire JA, Pandita TK, Bradley G, Kuba RM, Zhang H, Gulyas S, Hill RP, Nolan GP, Benchimol S . 1999 Mol. Cell. Biol. 19: 2373–2379
Veldman T, Horikawa I, Barrett JC, Schlegel R . 2001 J. Virol. 75: 4467–4472
Wang J, Xie LY, Allan S, Beach D, Hannon GJ . 1998 Genes Dev. 12: 1769–1774
Wang Z, Kyo S, Takakura M, Tanaka M, Yatabe N, Maida Y, Fujiwara M, Hayakawa J, Ohmichi M, Koike K, Inoue M . 2000 Cancer Res. 60: 5376–5381
Wick M, Zubov D, Hagen G . 1999 Gene 232: 97–106
Wright WE, Piatyszek MA, Rainey WE, Byrd W, Shay JW . 1996 Dev. Genet. 18: 173–179
Wu KJ, Grandori C, Amacker M, Simon-Vermot N, Polack A, Lingner J, Dalla-Favera R . 1999 Nat. Genet. 21: 220–224
Xu D, Popov N, Hou M, Wang Q, Bjorkholm M, Gruber A, Menkel AR, Henriksson M . 2001 Proc. Natl. Acad. Sci. USA 98: 3826–3831
Xu D, Wang Q, Gruber A, Bjorkholm M, Chen Z, Zaid A, Selivanova G, Peterson C, Wiman KG, Pisa P . 2000 Oncogene 19: 5123–5133
Yan P, Coindre JM, Benhattar J, Bosman FT, Guillou L . 1999 Cancer Res. 59: 3166–3170
Yan P, Saraga EP, Bouzourene H, Bosman FT, Benhattar J . 2001 J. Pathol. 193: 21–26
Yang J, Chang E, Cherry AM, Bangs CD, Oei Y, Bodnar A, Bronstein A, Chiu CP, Herron GS . 1999a J. Biol. Chem. 274: 26141–26148
Yang X, Tahin Q, Hu YF, Russo IH, Balsara BR, Mihaila D, Slater C, Barrett JC, Russo J . 1999b Int. J. Oncol. 15: 629–638
Yu CC, Lo SC, Wang TC . 2001 Biochem. J. 355: 459–464
Zhang A, Zheng C, Lindvall C, Hou M, Ekedahl J, Lewensohn R, Yan Z, Yang X, Henriksson M, Blennow E, Nordenskjold M, Zetterberg A, Bjorkholm M, Gruber A, Xu D . 2000 Cancer Res. 60: 6230–6235
We thank Marc van de Wetering and Hans Clevers for providing us with RNA from colon carcinoma lines carrying tetracycline inducible dominant negative TCF variants, and allowing us to quote their unpublished results. One of us is much indebted to the medical and nursing staff of the Otorhinolaryngology unit of the Lausanne University Hospital and to Dr F Viani for their help in making it possible for him to get on with his contribution to the review. Work in the authors' laboratories is supported by the Swiss National Science Foundation, the Human Frontiers Science Program, the fifth framework program of the European Union (via the Bundesamt für Bildung und Wissenschaft, Bern), and Krebsforschung Schweiz and the Roche Research foundation.
About this article
Cite this article
Ducrest, AL., Szutorisz, H., Lingner, J. et al. Regulation of the human telomerase reverse transcriptase gene. Oncogene 21, 541–552 (2002). https://doi.org/10.1038/sj.onc.1205081
TERT promoter methylation and protein expression as predictive biomarkers for recurrence risk in patients with serous borderline ovarian tumours
External environmental agents influence telomere length and telomerase activity by modulating internal cellular processes: Implications in human aging
Environmental Toxicology and Pharmacology (2021)
Clinical and diffusion parameters may noninvasively predict TERT promoter mutation status in grade II meningiomas
Journal of Neuroradiology (2021)
Human telomerase reverse transcriptase depletion potentiates the growth-inhibitory activity of imatinib in chronic myeloid leukemia stem cells
Cancer Letters (2020)
Applied Immunohistochemistry & Molecular Morphology (2020)