A dyskerin motif reactivates telomerase activity in X-linked dyskeratosis congenita and in telomerase-deficient human cells

Telomeres consist of TTAGGG repeats,¹  which shorten in each cell cycle due to the inability of DNA polymerases to replicate the ends of linear DNA molecules. This telomere erosion is counteracted by telomerase, a protein-RNA molecule that mediates the synthesis of the telomeric repeats from a template encoded by telomere RNA (TR).^2,3 

Telomere shortening, or loss of secondary structure, causes cells to enter senescence and/or apoptosis. These 2 processes act as biologic checkpoints to prevent uncontrolled cell division and genetic instability.^4,5  Defects in telomere length have been implicated in the pathology of several diseases, such as premature aging syndromes and cancer.⁶  Active telomerase complexes comprise hTERT, hTR, and Dyskerin.⁷  Telomerase expression is repressed in most postnatal somatic cells but is maintained in the majority of human cancers, thus contributing to the immortal phenotype by ensuring telomere integrity.⁶  Telomerase activity is regulated, mainly, at hTERT transcription level⁸  and, in humans, is controlled by a 5′ regulatory region of the hTERT gene, which is required for maximal activity and contains a myc binding site (E-box).^8,9  c-MYC transcription factor stimulates hTERT expression, whereas enforced expression of mad1 represses hTERT.^10,11  Transcriptional regulation of c-myc involves multiple promoters, the predominant ones being P1 and P2.¹²  The nuclease hypersensitive element-III (NHEIII) of the c-MYC promoter controls 85% to 90% of c-MYC transcription and was originally identified as a major DNAseI hypersensitivity site.¹³  NHEIII contains a purine-rich encoding strand capable of establishing an equilibrium between a duplex-helix structure and non–B-forms of DNA.¹⁴  The DNA-purine–rich strand can form 2 different intramolecular G-quadruplex structures,¹⁵  which, since they function as transcriptional repressor elements, have been targeted for different anticancer therapies.

One of the premature aging syndromes associated with telomere erosion is dyskeratosis congenita (DC). It is a rare, inherited disorder characterized by bone marrow failure and increased susceptibility to cancer,^16,–18  the main cause of mortality in these patients. The X-linked form of DC (X-DC) is caused, mainly, by point mutations in the DKC gene.^19,20  This gene codes for dyskerin, a 58 kDa protein with a putative pseudouridine synthase domain, a component of the telomerase complex. Dyskerin binds to the H/ACA box present in SnoRNAs (small nucleolar RNAs) and TR in vertebrates.²¹  SnoRNAs act as a guide for dyskerin-mediated rRNA pseudouridylation. H/ACA RNAs contain 2 hairpins of variable length separated by a single-stranded region that includes a conserved H box, and followed by a short tail containing an ACA trinucleotide.²²  Three different proteins (Nap2, Gar1, and NOP10) associate with DKC at H/ACA SnoRNAs.^3,22,–24  DKC, Nap2, and NOP10 are tightly associated and are involved in maintaining the stability of the ribonucleoprotein complex, whereas interaction with Gar 1 is labile and transient.^25,,–28  Therefore, dyskerin is required for telomerase function and is essential for the nuclear accumulation of hTR^{7,24,,,,–29}  and rRNA pseudouridylation.³⁰  Chromosomal instability in patients with DC increases with age as a consequence of decreased telomerase activity and incorrect maintenance of telomere structure. In X-DC lymphoblasts and fibroblasts, telomerase activity and hTR levels are reduced²¹  and telomeres are shorter than in nonaffected cells. The telomerase defect can be overcome by forced expression of hTERT and hTR.^21,31  Other forms of DC result from mutations or deletions in hTR, resulting in accelerated telomere shortening and premature death. This defect can only be overcome by hTR re-expression.³¹ 

In the present study we present data on a genetic suppressor element (GSE) named 24-2, coding for the pseudouridine synthase domain of DKC. Expression of this domain can increase survival against cisplatin and telomerase inhibitors in human cells. GSE24-2 activates c-myc promoter through NHEIII and, consequently, activates the hTERT promoter. There is a concomitant increase in telomerase activity in VA13 human cells, and in cells derived from patients with X-DC. Furthermore, expression of GSE24-2 rescues X-DC cells from premature senescence. This suggests that expression of GSE24-2 could be a new therapeutic approach for telomerase-deficient syndromes such as X-DC.

All mice were housed under similar humane conditions in accordance with the regulations of the Spanish Government concerning experimental animal research.

GSE24-2, DKC, DKC5′, motif I, motif II, and the CBF5 region, homologous to GSE24-2, were cloned in pLNCX (BD Biosciences, San Jose, CA). The hTERT promoter construct was a gift from Dr T. Kim (Institute for Molecular Biology and Genetics, Seoul National University, Seoul, Korea)³² ; the hTR promoter construct was a gift from Dr N. Keith (Centre for Oncology and Applied Pharmacology, University of Glasgow, United Kingdom)⁹ ; those of the c-MYC (px3.2) promoter were from Dr A. Aranda (Instituto de Investigaciones Biomedicas, Madrid, Spain); and the Myc/mad hybrid construct, pECL and pECLc-myc, were from Dr J. León (Departamento de Biologia Molecular, Universidad de Cantabria, Santander, Spain). BABE-TERT plasmid was obtained from K. Collins (Department of Molecular and Cell Biology, University of California at Berkeley, CA).³¹  PGATEV plasmid³³  was obtained from Dr G. Montoya (Centro Nacional de Investigaciones Oncológicas, Madrid, Spain). PGATEV-24-2 was obtained by subcloning the 24-2 fragment into the NdeI/XhoI sites of the pGATEV plasmid. The PCDNA3-9E10-24-2 sequence was obtained by subcloning the 24-4 fragment between the EcoR1/XbaI sites of the vector containing the 9E10-myc–recognized epitope. 293T (American Type Culture Collection, Manassas, VA) and Phoenix cells were maintained in Dulbecco modified Eagle medium (DMEM) 10% fetal bovine serum (FBS). The VA13 cell line was obtained from Dr M. Serrano. X-DC cells were obtained from the Coriell Cell Repository and maintained in RPMI 20% FBS. The X-DC lymphoblast family (DC, DC1, DC2, and DC3) has been clinically described^34,35  and the dyskerin gene has the single amino-acid substitution T66A. X-DC GMO1787 dermal fibroblasts coding for a leucine 3–lacking dyskerin protein³¹  were cultured in DMEM supplemented with 10% fetal calf serum (FCS). Murine embryonic fibroblasts (MEFs) were cultured in DMEM supplemented with 10% FBS and nonessential amino acids (Invitrogen, Carlsbad, CA). For stable transfection, 293T cells were cotransfected with 10 μg/million cells GSE24-2 or pLNCX and 1 μg p-BabePUR plasmid (Clontech, Mountain View, CA) and selected for puromycin resistance 24 hours after transfection. Pam 212 cells were transfected with pECL vector, pECLc-myc, or GSE24-2 and were selected for G418 resistance. For infection of X-DC fibroblasts, Phoenix cells were transfected with the pLNCX-derived vectors using calcium phosphate. The viruses were collected 48 hours after transfection and used to infect X-DC fibroblasts over 48 hours. Stable cells were selected with G-418 (Invitrogen).

The GSE library was constructed essentially as described by Roninson et al.³⁶ 

Cisplatin, telomerase inhibitor I, and SP60125 were purchased from Calbiochem (San Diego, CA). Tumor necrosis factor α (TNFα) was purchased from Upstate Biotechnology (Charlottesville, VA).

E coli DH5α cells were transformed with pGATEV GSE24-2 and lysates were prepared as described.³³  The fusion protein was purified with glutation-sepharose and purity was analyzed by gel electrophoresis. GSE24-2 was obtained by TEV protease digestion according to the manufacturer's recommendations. Typically, over 90% of the fusion protein was cleaved, as determined by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE). The protein was passed twice over a 5 mL Hi-Trap Ni-NTA column to remove the polyhistidine tags, uncleaved protein, tobacco etch virus (TEV) protease, and impurities.

Mutation of single guanines in the NHEIII region of the c-MYC promoter (px3.2) was performed using the Quickchange X-L site-directed mutagenesis kit (Stratagene, Santa Clara, CA) according to the manufacturer's protocol.

Telomerase activity was measured using the TRAPeze³⁷  telomerase detection kit (Serologicals, Norcross, GA) according to the manufacturer's recommendations. The telomeric repeat amplification protocol (TRAP) assay was performed following a titration of each extract and normalized for total protein concentration.³⁶  Where indicated, a telomere-specific probe was used for the TRAP assay.³⁸  The in vitro platinum-exposure assays had cisplatin added directly to the TRAP reaction.

Cells were lysed as previously described³⁹  and 20 μg of protein analyzed in 10% SDS-PAGE. Antibodies used were: anti-pJNK (V7391; Promega, Madison, WI); anti-JNK1 (C-17; Santa Cruz Biotechnologies, Santa Cruz, CA); anti p-P38 (Cell Signaling Technology, Danvers, MA); 9-E10 (A14; Santa Cruz Biotechnologies); and anti-Flag (Invitrogen, Carlsbad, CA).

Total cellular RNA was extracted using Trizol (Invitrogen) according to the manufacturer's instructions. In each reaction, 2 μg total RNA was reverse transcribed into cDNA using M-Mlv reverse transcriptase (Promega). For a list of oligonucleotides used, see Figure S1 (available on the Blood website; see the Supplemental Materials link at the top of the online article).

293T cells were transiently transfected using the calcium phosphate method as previously described.³⁹  The total amount of DNA was kept constant at 10 μg per million cells. X-DC cells were transiently transfected by electroporation of 3 μg pLNCX-GSE24-2 per million cells. VA13 cells and MEFs were transfected with 16 μg DNA/million cells using lipofectamine plus (Invitrogen), according to the manufacturer's instructions. Protein extracts were prepared and luciferase determined using a commercial kit (Promega). Each assay was performed in triplicate, and repeated in 3 different experiments. Transfection efficiencies were corrected by cotransfection of p-CMV-Renilla.

293T cells were transfected with PCDNA3-E1024-2 or an empty vector and stable cells were obtained by G-418 selection. These cells expressed an E10-myc epitope fusion GSE24-2 protein. Nuclear protein extraction⁴⁰  and binding reactions were performed essentially as described previously.⁴¹  Oligonucleotides TEL1, containing the telomeric sequence, and CXext, which hybridized to the TEL1 overhang (a negative control),⁴¹  were labeled at the 5′ end with T4 polynucleotide kinase (New England Biolabs, Ipswich, MA) and γ³²P ATP (Amersham Bioscences, Orsay, France). Specific competition of GSE24-2 binding was performed with the tag-antibody 9E10. DNA-bound complexes were separated in nondenaturing polyacrylamide gels in 0.5% tris borate EDTA buffer.

Cell growth and viability were determined using a crystal violet staining method, as previously described.³⁹  DKC fibroblasts were pooled after infection and selection for G-418 resistance, and grown on bulk culture with the population doubling (PLD) count initiated at the first split of postselection culture.

A GSE approach was designed to isolate cDNA sequences conferring survival capacity to cells treated with cisplatin. A cDNA library normalized for low-abundance mRNAs³⁶  was prepared from human placenta, cloned into a retroviral vector, and used to infect Pam212 cells. Cells growing at a cisplatin concentration of 3 μg/mL were selected; 100 resistant clones were isolated and 5 containing a 165-bp fragment of dyskerin were termed GSE24-2. The DNA sequence of GSE24-2 was localized at the N-terminal sequence of dyskerin (Figure 1A). Since the fragment isolated from the library contained 3 different reading frames,³⁶  we subcloned the open reading frame corresponding to dyskerin in the GSE24-2 into the pLNCX vector. In human 293T cells expressing GSE24-2, the 50% inhibitory concentration (IC50) to cisplatin was close to one order of magnitude higher than control cells (Figure 1B). The activity of this plasmid was equivalent to the DNA fragment isolated in the original screening, and it strongly indicated that cisplatin resistance was indeed induced by an internal sequence of dyskerin. Hence, we used this construct for the experiments presented in this report. We had previously shown that JNK/p38 signaling pathway activation is required for apoptosis in cisplatin-treated cells.³⁹  We studied the response in GSE24-2 and control cell lines and found that GSE24-2 cells required higher doses of cisplatin for JNK and p38 activation (Figure 1C). As activation of both kinases is related to DNA damage response,³⁹  the results indicated that GSE24-2–expressing cells were less responsive to DNA damage induced by cisplatin, and that the increased survival against cisplatin correlated with biochemical parameters related to cell death.³⁹  These results are specific for cisplatin since the survival of GSE24-2 cells against Taxol (Sigma-Aldrich, St Louis, MO) was equivalent (data not shown) and, with adriamycin, GSE24-2–expressing cells were more sensitive than control cells (data not shown). This correlated with an increased phosphorylation of p38 in GSE24-2–treated cells (data not shown).

To determine whether cisplatin resistance was due specifically to expression of the 165-bp fragment, we cloned the full-length cDNA for dkc(DKC) as well as the first 495 nucleotides that included the GSE24-2 sequence (DKC 5′), into pLNCX (Figure 1A). Cells expressing GSE24-2 showed less sensitivity to cisplatin than the other cell lines (Figure 1B). Thus, the activity that induced resistance to cisplatin was limited to the GSE24-2 fragment, which includes the pseudouridine synthase domain of DKC,^19,,–22  a domain that is highly conserved in all eukaryotes (Figure 1D).

The TTAGGG telomeric sequence is a good target for cisplatin. Intramolecular quadruplexes formed at the telomere consist of 4 consecutive human telomeric DNA repeats.⁶  These structures are cross-linked by cisplatin. The result is inhibition of telomerase and shortening of the telomeres.⁴²  Therefore, since DKC is one of the components of the telomerase complex,⁷  we studied the effect of cisplatin and the expression of GSE24-2 on telomerase activity using the TRAP assay.³⁷  Control cells (pLNCX) as well as GSE24-2 cells were pretreated for 3 days with a low cisplatin concentration of 0.5 μg/mL. At this dose, replication takes place without telomere formation.^42,43  Cells were then treated with the concentration used for GSE selection (3 μg/mL) for 3 or 7 days. At 7 days (Figure 2A), we observed inhibition of telomerase activity in pLNCX cells, as has been described in the literature.^42,43  However, GSE24-2 cells showed higher telomerase activity even after 7 days of treatment (Figure 2A dots). Indeed, the level of activity in control cell extracts (lane 1) matched the level of activity in GSE24-2 cell extracts treated for 3 or 7 days (Figure 2A lanes 13-16). In an in vitro platinum dose-response assay, the level of telomerase activity in GSE24-2 cell extracts, following incubation with 5 μg/mL or 20 μg/mL cisplatin (Figure 2B lanes 7-9), matched the level of activity in control cell extracts (drug-free or with 5 μg/mL of the drug; Figure 2B lanes 1,2). Therefore, cells expressing GSE24-2 required higher doses of cisplatin and longer exposure times to produce an inhibition of telomerase activity. Moreover, since the TRAP probe used in the in vitro platinum exposure assay was telomeric,^3,37  the activity of GSE24-2 was also able to protect telomeres from cisplatin crosslinking. Telomerase activity inhibition by cisplatin was also assayed in vitro using cell extracts from 293T cells and a telomeric probe, treated with cisplatin (Figure 2C). In this assay, if GSE24-2 peptide was added, a dose-dependent protection of telomerase activity of cell extracts (Figure 2C lanes 3-5) against cisplatin inhibition was achieved. In contrast, heat-inactivated GSE24-2 peptide or a nonrelated peptide was not able to protect against cisplatin-induced telomerase inhibition (Figure 2C lanes 6,7). This is a further indication that inhibition was, indeed, on telomerase activity and that GSE24-2 peptide was able to inhibit cisplatin-induced telomere crosslinking in vitro. We analyzed the possible binding of GSE24-2 to the telomeric probe using gel mobility shift assays and a short hairpin oligonucleotide (TEL1), which reconstitutes the double-stranded telomere with a short 3′ overhang.⁴¹  One specific GSE24-2/TEL1 complex was detected (Figure 2D). CXext, which hybridizes to the single-stranded extension of TEL1, inhibits complex formation. Altogether these results indicate that GSE24-2 was able to bind telomeric sequences; the lack of inhibition of telomerase activity by cisplatin could be due in part to a “protection effect” of GSE24-2 on the telomeric probe.

Other compounds that stabilize telomeric G-rich single-strand overhangs and block telomere replication are considered antitumor agents.^42,,,–46  The telomerase inhibitor I (TI1; 2,6-diaminoanthraquinone),⁴⁷  a compound which stabilizes the G-quadruplex structures, was less toxic in GSE24-2 cells than in control cells (data not shown), indicating that the product of GSE24-2 expression would interfere with the capacity of TI1 to stabilize the G-quadruplexes.⁴⁸  Accordingly, cells expressing GSE24-2 required longer periods of TI1 treatment to inhibit telomerase activity (Figure 3A lanes 5,6 and 14,15).

Resistance to G-quadruplex ligands has been associated with telomerase overexpression¹⁰  in human cells. We investigated whether the resistance to cisplatin and TI1 induced by GSE24-2 could be due to increased expression of some of the telomerase complex components, resulting in higher levels of telomerase activity. The effect of cisplatin on the expression of hTR, hTERT, and dyskerin was determined by semiquantitative PCR. Cisplatin induced a decrease in hTERT mRNA at 3 days of treatment in control cells as previously described.^49,–51  However, 7 days of treatment were required for a comparable decrease in GSE24-2 cells (Figure 3B). We did not observe a significant effect of cisplatin on hTR RNA or dkc mRNA levels (Figure 3B). The effect of TI1 treatment on hTERT and hTR RNA levels was also tested (Figure 3C). hTERT mRNA decreases in pLNCX cells after 3 days of treatment whereas in GSE24-2 cells this decrease was observed only after 7 days of treatment. No significant differences were observed in hTR levels (Figure 3C). These results suggest that the differences observed in telomerase activity following cisplatin or TI1 treatment were due, at least in part, to the ability of GSE24-2 cells to maintain higher hTERT RNA levels than control cells. In agreement with this hypothesis, 293T cells that overexpressed hTERT (Figure 3D) were more resistant to cisplatin (Figure 3E) than control cells.

Since cisplatin induces a decrease in hTERT mRNA and the expression of GSE24-2 appears to compensate for this effect, we investigated whether this could be due to changes in hTERT transcription.⁵²  We assessed the activity of the hTERT promoter in GSE24-2 cells using a luciferase reporter construct containing 3402 bp of the promoter.³²  Expression of GSE24-2, but not DKC, was able to stimulate hTERT promoter activity in transient transfection assays (Figure 4A). Further, cisplatin stimulated higher promoter activity in GSE24-2 than in control cells (Figure 4B). These results indicate that expression of GSE24-2 was able to stimulate transcription even in the presence of cisplatin, resulting in increased levels of hTERT mRNA (Figure 3B). Further, we were able to detect stimulation of hTERT promoter in MEF cells transfected with the GSE24-2 plasmid (data not shown). We did not observe any effect of GSE24-2 expression on hTR promoter activity⁹  using either DKC or GSE24-2 cells (Figure 4C). Conversely, treatment with the JNK inhibitor SP60125 was able to activate this promoter (data not shown), as has been previously described.⁹  These results indicate that the activity of GSE24-2 was specific for hTERT expression.

Dyskerin was originally identified as the human cbf5 homologue⁵³  showing 85% identity to S cerevisiae Cbf5 gene at the protein level.⁵⁴  Therefore, we cloned the yeast S cerevisiae cbf5 gene sequence equivalent to GSE24-2 (Figure 1D) into the pLNCX vector. Expression of this construct also increased hTERT(Figure 4A), but not the hTR promoter activity (Figure 4C). This indicates a high degree of functional conservation in this DKC region.

The pseudouridine synthase TRUB domain of DKC comprises 2 structural subdomains, motif I and motif II,⁵³  which exactly match the GSE24-2 encoded region (Figure 1D). These 2 domains are critical for maintaining the global protein structure of DKC. Additionally, motif II contains at least one residue (asp125) essential for enzymatic activity. We assayed the effect of both individual motifs on hTERT promoter activation. Only motif I significantly increased the activity of the hTERT promoter, indicating that it contains the minimal sequence required for maximal activation (Figure 4A).

The hTERT core promoter contains 2 E-boxes, which can bind myc/max heterodimers.^32,55  A hybrid molecule containing the DNA-binding domain of c-myc and the transactivation domain of mad is able to bind E-boxes, thus inhibiting c-myc–dependent transcription. Expression of mad/myc was able to inhibit the activity of the hTERT promoter in a dose-dependent manner in control and DKC cells (Figure 4D). Moreover, the expression of myc/mad was able to abolish transcription activation mediated via the GSE in GSE24-2 cells. This indicated that activation of hTERT transcription induced by GSE24-2 was c-myc dependent (Figure 4D). Inhibition was specific since transfection of the myc/mad construct together with an NFκB-dependent promoter (HIVLuc) did not affect transcription following stimulation with TNFα (Figure 4E).

A reporter vector containing 3.2 kb of c-myc gene (px3.2myc), including the P0, P1, and P2 promoters linked to the luciferase gene (Figure 5A), was transfected in control and GSE24-2 cells. A 3-fold increase in activity was observed in the GSE24-2 cells (Figure 5B). A similar result was obtained in 293T cells transfected with either pLNCX vector or the plasmid expressing GSE24-2 together with the px3.2myc reporter (data not shown). Three deletion mutants of the c-myc promoter were used to define the region necessary for GSE24-2 activation (Figure 5A). Only those regions containing the distal P1 and P0 promoters, including NHEIII, were able to support transcription in GSE24-2, but not in control cells (Figure 5B). This indicated that the proximal P2 promoter was not involved in GSE activation.

Several NHEIII mutants were obtained by substituting guanine residues at the purine-rich region involved in maintaining the secondary structure of G-quadruplex (Pu27). We mutated G12A at the second G-quartet group, G17A at the second G-triplet and, finally, 2 consecutive guanines (G26A/G27A) at the end of the purine-rich region (Figure 5C). The results indicated that only the wild-type (WT) px3.2 myc promoter region, but none of the mutants of the Pu27 region, is activated by GSE24-2 (Figure 5D). Taken together, these results suggest that GSE24-2 is able to induce a modification of the secondary structure of the Pu27 region into an active conformation enabling c-myc transcription.

We further explored whether these changes in promoter activity resulted in c-myc expression. MEF cells were transiently transfected with the empty vector, DKC, or GSE24-2–expressing vectors, and protein and mRNA levels of c-myc were analyzed (Figure 5E,F). As expected, only expression of GSE24-2 was able to cause increased c-MYC mRNA and protein expression. One of the undesirable effects of GSE24-2 expression could be that an increase in c-MYC expression would induce cell transformation. To evaluate this possibility, nude mice were inoculated with Pam212 cells expressing empty vector, human c-myc, or GSE24-2. Only animals inoculated with c-myc developed tumors during the first 10 days, whereas GSE24-2 cells did not develop tumors even after 3 months (data not shown), indicating that expression of GSE24-2 is not tumorigenic.

X-DC cells have low levels of telomerase RNA, reduced levels of telomerase activity, and shorter telomeres than normal cells.²¹  Hence, we evaluated the effect of GSE24-2 expression on telomerase activity of cells derived from patients with X-DC. Commercially available lymphoblasts from a carrier mother (DC, normal dkc allele) and 3 of her clinically affected sons (DC1, DC2, and DC3, mutant dkc allele) were used. These cells are not immortalized and eventually enter senescence in culture. Expression of GSE24-2 increased telomerase activity in all these cell lines (Figure 6A). Interestingly, we did not observe an increase in telomerase activity upon expression of full-length DKC (Figure 6B), as previously reported.²¹  An increase in hTERT promoter activity was also observed upon GSE24-2 transfection (data not shown). Compared with the healthy control (carrier) mother, hTR levels are low in DC cells expressing a mutant dkc gene^21,56  (Figure 6C). RNA levels of hTERT and hTR increased in cells transfected with GSE24-2 (Figure 6C), suggesting that recovery of telomerase activity upon GSE24-2 transfection is due to an increase in both molecules. Primary dermal fibroblasts from patients with X-DC became completely senescent within 3 to 4 weeks of continuous culture, whereas normal cells proliferate up to 2 months before reaching senescence.²¹  X-DC cells expressing hTERT acquired extended life span and grew continuously for several months.²¹  Expression of GSE24-2 in GMO1787 X-DC cells was able to induce an increase in telomerase activity (Figure 6D). We also found an increase in hTR and hTERT RNA levels in X-DC fibroblasts expressing GSE24-2 (Figure 6E), and these cells were able to grow continuously (Figure 6F); the results resembled those obtained after hTERT expression.³¹  Interestingly, in the telomerase-deficient cell line VA13 (wild-type dkc), which lacks hTERT and hTR⁵⁷  and maintain telomeres by ALT mechanism, expression of GSE24-2 was able to recover telomerase activity (Figure 6G), hTERT RNA, and hTR (Figure 6H).

Taken together, these results indicated that expression of GSE24-2 was able to trigger a recovery of telomerase activity by up-regulating hTERT and hTR RNA levels in both types of cell lines; those expressing a mutant DKC and those with a wild-type DKC allele and low telomerase activity. This increase in telomerase activity resulted in rescue of senescence in X-DC fibroblasts.

Controlling telomere length is of considerable relevance in cancer development and aging.⁶  The findings reported in this study provide the basis for a therapeutic tool that can be used for reactivation of telomerase activity in telomerase-defective syndromes characterized by a faster rate of telomere loss and genomic instability. The therapeutic application can also be for premature aging diseases, inducing a rescue from senescence.

GSEs are useful tools for functional screening of proteins and protein domains.³⁶  We have shown that expression of a dyskerin internal domain is able to induce cisplatin resistance. This is specific for lesions induced by cisplatin since it does not confer resistance to other DNA-damaging agents such as adriamycin. This difference in behavior might be related to the damage that cisplatin induces to the telomeres, given the reactivity of cisplatin with guanines in the DNA. The nonrepaired cisplatin adducts on the telomere repeats will block DNA replication,⁴²  thus inducing apoptosis. A significant shortening of the telomeres has been reported at cisplatin doses used for the in vitro selection of the GSEs.⁴²  Our results indicate that GSE24-2 is able to protect telomeres from cisplatin crosslinks in vitro and up-regulate telomerase activity in vivo. Besides, GSE24-2 cells are also more resistant to TI1 and this would support the concept that GSE could have some effect on G-triplets at telomeres.⁵⁸  Support for this hypothesis is the down-regulation of telomerase activity observed in response to cisplatin in the cisplatin-sensitive ovarian carcinoma cell line A2780, but not in the cisplatin-resistant cell line A2780CP.⁵⁹  Conversely, telomerase activities in both cell lines were reduced when cells were treated with adriamycin. This indicates different effects on telomerase activity of the 2 anticancer drugs.⁵⁹  In addition, GSE24-2 rescues the cisplatin stabilization of the G-quadruplexes formed at the NHEIII site in the c-myc promoter.

Regulation of TERT mRNA expression appears to be the most important limiting step for telomerase activation⁶  since hTERT is not expressed in cells approaching senescence. GSE24-2 is involved in transcriptional control of hTERT. There is a good correlation between the ability of GSE24-2 to induce cell survival (and to protect telomerase activity against cisplatin) and hTERT promoter activation. This suggests that an increase in telomerase expression would be, at least in part, responsible for increased survival against cisplatin of GSE24-2–expressing cells. Consequently, overexpression of hTERT in 293T cells induced survival against cisplatin. This is also supported by the observation that activation of the TERT promoter in GSE24-2 cells is higher in the presence of cisplatin than under basal conditions (2.5-fold vs 5-fold). hTERT promoter is also activated in cells transfected with the GSE24-2 equivalent fragment from the yeast cbf5 gene,^53,54  which implies an evolutionary conserved function of this domain, relevant for hTERT expression.

It is very likely that GSE24-2 activates hTERT transcription in modulating c-myc expression since the myc/mad hybrid molecule totally abolished hTERT transcriptional activation. The GSE24-2 target region appears to be a polypurine sequence (Pu27) also found in other promoters such as CCR5 and PDGFA, and is often associated with unusual DNA structures.⁶⁰  Mutational analysis suggests that GSE24-2 requires an intact Pu27 to stimulate c-myc transcription. Therefore, alteration of the secondary structure at the G-repeats of the NHEIII site will impair GSE24-2 activity. There are a few known proteins that interact and activate NHE sites. One of them is NM23-H2/NDP kinase B, a sequence-specific DNA-binding protein with affinity for the NHE site at the c-myc promoter. It also activates the expression of other genes such as myeloperoxidase, CD11b, and CCR5, and represses the expression of others such as PDGFA.^60,61  Interestingly, this protein can bind single-strand G-enriched sequences in vitro,^60,61  including those at the NHE-c-myc promoter and human telomeres. Therefore, a possible mechanism by which GSE24-2 can activate transcription through the NHE is by association with proteins such as NM23H2, which facilitate their interaction with DNA and thereby increase transcription.

Our results obtained in X-DC and VA13 cells indicated that GSE24-2 could increase telomerase activity through a combination of 2 different mechanisms: (1) activating transcription of hTERT, and (2) increasing hTR levels by stabilization of the RNA, in the absence of promoter activation. In addition, GSE24-2 might facilitate the assembly of other proteins such as hNaf1 or Nop10 within the H/ACA box, even in the presence of defective DKC.²⁸  We had not observed any increase in hTR in 293T cells, probably because active telomerase complexes are already present in these cells. Conversely, expression of full-length dyskerin does not result in hTR accumulation in cells from patients with X-DC, as has been described previously,³¹  probably because the increase in DKC levels is potentially very disruptive for the biogenesis and function of H/ACA motif RNPs.

Independently of the possible mechanism involved, the present results open up the possibility of treatment of X-DC cells with the small GSE24-2 gene fragment, or the encoded peptide. The outcome would be to recover telomerase activity in X-DC cells and maintain hematopoietic renewal, the progressive loss of which is fatal for patients with X-DC. Our results indicate that GSE24-2 expression does not induce tumorigenicity and would not result in aberrant cell growth. Indeed, Wong and Collins³¹  recently reported that the telomerase defect in patients with X-DC suffering premature cell senescence arise from reduced accumulation of hTR. Our results demonstrate that expression of GSE24-2 is able to increase hTR and hTERT levels in X-DC cells, and recover the capacity to divide. Further, GSE24-2 increased hTERT and hTR expression in VA13 cells. Therefore, the possibility arises that it could be used to rescue telomerase activity in different cell types that have telomerase defects, including cells expressing normal DKC such as VA13, and those from other human syndromes such as autosomal dominant DC or aplastic anemia. Temporary telomerase reactivation could prevent critical telomere shortening and associated pathologies in premature aging syndromes characterized by an increased rate of telomere loss and genomic instability. Telomerase reactivation could be used therapeutically to target those dividing-cell types that maintain organ homeostasis. Telomerase activation therapy could also be used to increase survival of stem-cell populations in patients receiving human stem cell (HSC) transplantation treatment since, even if these cells express telomerase, they are not able to maintain telomere length over a protracted period of time.⁶ 

We gratefully acknowledge M. Blasco and M. Serrano for helpful suggestions; M. Bessler and P. Mason for suggesting some experiments; and N. Keith, A. Aranda, J. León, K. Collins, T. K. Kim, and G. Montoya for providing plasmids used in this work. We also acknowledge J. Perez for the artwork.

This work was supported by grants 04/1164 and 05/1305 from Fondo de Investigación Sauitaire (FIS) and Biologia Fundamental BFU-05-0138. R.M.-P. is supported by Fundación Genoma and grant C03/10 provided by Red Temàtica de Investigación de Centros de Cáncer (RTICC). I.S.-P. and J.-R.M. are Ramón y Cajal program researchers funded by Ministerio de Educación y Ciencia (MEC).

Contribution: I.S.-P. and R.M.-P. performed the greater part of the experimental work and contributed equally to the manuscript preparation; J.-R.M. implemented the initial experiments; L.S. provided assistance on specific techniques and experimental design; R.P. was responsible for experimental design and drafting the manuscript.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Rosario Perona, Instituto de Investigaciones Biomédicas, C/Arturo Duperier, 4, 28029 Madrid, Spain; e-mail: rperona@iib.uam.es.