Loss of telomeric repeats has been causally linked to replicative senescence and aging in human cells. In contrast to normal somatic cells, which are telomerase-negative, hematopoietic stem cells have low levels of telomerase, which can be transiently upregulated upon cytokine stimulation. To examine whether ectopic expression of telomerase can overcome telomere erosion in hematopoietic progenitor cells, we overexpressed telomerase in CD34+ and AC133+ cord blood (CB) cells using retroviral vectors containing hTERT, the catalytic component of telomerase. Although the hTERT-transduced CB cells exhibited significantly elevated telomerase activity (approximately 10-fold), the mean telomere length was only increased up to 600 bp, which was in contrast to hTERT-transduced fibroblast cells gaining more than 2-kb telomeric repeats. Moreover, ectopic telomerase activity did not prevent overall telomere shortening, which was in the range of 1.3 kb in serum-free expansion culture. We also blocked endogenous telomerase activity by ectopic expression of dominant-negative hTERT. Whereas CB cells with absent telomerase activity showed reduced absolute numbers of colony-forming cells, we observed increased rates only for burst-forming units erythroid when the enzyme was overexpressed. These results suggest that telomere shortening in human hematopoietic progenitor cells cannot be compensated by increased levels of telomerase alone and is likely to be dependent on other factors, such as telomere binding proteins. Furthermore, telomerase function seems to be directly associated with the proliferative capacity of stem cells and may exert an additional role in lineage differentiation.
Hematopoietic stem cells (HSCs) have a remarkable regenerative potential to replace most mature blood cells continuously throughout life. However, serial bone marrow transplantation experiments in mice associated with shortening of telomere repeats provide strong evidence that HSCs are not immortal cells but have a limited replication potential [1, 2]. Telomeres are the molecular caps at the ends of chromosomes that are composed of repetitive TTAGGG sequences and associated proteins . Telomeres stabilize and protect chromosome ends from end-to-end fusions, recombination, and degradation. The maintenance of telomeres throughout many cycles of cell division requires the enzyme telomerase, which consists of two core components, the RNA-subunit, hTR, containing the template, and a catalytic protein subunit, hTERT . In general, telomerase activity is absent in most somatic human cells, and programmed shortening of telomeres has been observed in dividing cells and with aging [5–8]. In contrast, most tumor and embryonic stem cells circumvent replicative senescence by expression of hTERT, which leads to stabilization of telomeres and acquisition of an immortal phenotype [9–11].
Unlike normal somatic cells, telomerase activity has been detected at low levels in hematopoietic progenitor cells, which is upregulated in response to cytokine stimulation [12–14]. However, despite telomerase activity, telomere shortening is not prevented but delayed on proliferation of hematopoietic cells [14–16]. Replicative stress such as bone marrow reconstitution after autografting or allografting has been shown to result in accelerated telomere shortening . In addition, progressive telomere shortening has been reported in patients with bone marrow failure syndromes, such as aplastic anemia [18, 19].
Because telomerase plays a critical role in overcoming growth limitations attributable to telomere erosion, several normal cell types, such as fibroblasts, endothelial cells, and epithelial cells, have been immortalized with telomerase without signs of malignant transformation . Such attempts may open a range of opportunities for the use of telomerase-immortalized cells in research, tissue engineering, and treatment of age-related diseases. Particularly, activation of telomere maintenance mechanisms might pose a greater potential for adult stem cells similar to advantages of embryonic cells, which are highly telomerase positive and immortal [10, 11]. In this study, we manipulated the endogenous telomerase activity in hematopoietic progenitor cells by ectopic transfer of hTERT as well as by a dominant-negative mutant. As a result, telomere length dynamics were only minimally influenced despite striking differences in telomerase activity. This suggests that telomerase function is more tightly regulated in HSCs compared with various other cell types. Furthermore, we provide evidence that telomerase activity is directly associated with proliferative capacity and lineage differentiation.
Materials and Methods
Cell Lines, Purification, and Culture of CD34+ Cells
Cord blood (CB) samples from placental and umbilical tissues scheduled for discard were collected in heparin according to procedures approved by the institutional review board of the Freiburg University Hospital (provided by the Freiburg cord blood banking program). Mononuclear cells were obtained by Ficoll-Hypaque (Biochrom, Berlin) density centrifugation. CD34+ cells were purified by positive selection using a magnetic cell-sorting (MACS) progenitor enrichment kit according to the manufacturer's protocol (Miltenyi Biotec, Bergisch Gladbach, Germany). The purity of enriched progenitor cells was 92 ± 2% (n = 5).
Purified CD34+ cells were plated at a density of 5 × 104/ml medium using a 24-well plate (Becton, Dickinson, Heidelberg, Germany). Serum-free culture medium consisted of StemSpan (Stem Cell Technologies, Vancouver, Canada) supplemented with 40 μg/ml low-density lipoprotein (Sigma, Deisenhofen, Germany) and 300 ng/ml stem cell factor (SCF), 300 ng/ml Flt-3 L, 100 ng/ml interleukin (IL)-3, 10 ng/ml IL-6 (each provided by CellGenix, Freiburg, Germany), and 10 ng/ml G-CSF (Amgen, Thousand Oaks, CA). Cells were cultured at 37°C and 5% carbon dioxide.
Purity of enriched cells and phenotype of cultured cells were analyzed by a FACScalibur™ cytometer (Becton, Dickinson) using the following monoclonal antibodies (MABs): AC133-PE (Miltenyi Biotec), CD34-PE, and mouse IgG1-PE as an isotype control (all Becton, Dickinson). As primary diploid fibroblasts, the HK-1 cell line was used and cultured as described .
Retroviral Vectors and Infection Protocol
The full-length hTERT cDNA was inserted upstream of an internal ribosomal entry site (IRES) green fluorescent protein (GFP) cassette into a Moloney murine leukemia virus–based pBABE-vector (kindly provided by H. Vaziri and R. Weinberg, Massachusetts Institute of Technology, Boston). In addition, we used the recently described pOS vector  containing the hTERT-IRES-GFP construct as well as a dominant-negative mutant of hTERT, which was reported previously . Amphotropic virus supernatant was produced by calcium-phosphate transfection of phoenix ampho (ϕNX) packaging cells (kindly provided by G. Nolan, Stanford University, Stanford, CA) as previously described .
After prestimulation for 48 hours, CD34+ cells were infected with virus on retronectin-coated plates (4 μg/cm2; Takara, Otsu, Japan). Half of the medium was replaced with fresh retroviral supernatant of phoenix cells, which was repeated six times during 48 hours. Forty-eight hours after the last infection, transduced hematopoietic progenitor cells were sorted (day 6) to isolate GFP-positive fractions using a MoFlo (modular flow cytometer) high-speed cell sorter (Cytomation, Fort Collins, CO). Transduction efficiency was determined by flow-cytometric assessment of GFP expression, which was on average 33.7 ± 6.9% for the pBABE-hTERT-IRES-GFP–transduced and 67.1 ± 8.3% for GFP-only–transduced cells, respectively (n = 7). In a set of additional experiments, CD34+ CB cells were labeled with the MAB AC-133PE after the transduction period (n = 4). Essentially, an equivalent frequency of AC133+ cells was found in the GFP+ fraction in the range of 30% for the different vectors used (p = .8). Fluorescence-activated cell-sorted AC133+GFP+ cells were cultured for a total of 4 weeks as described above. At different time points of culture, cells were harvested, counted, and used for additional analysis and continuation of the culture.
Telomerase Repeat Amplification Protocol Assay
Telomerase activity of cell populations was determined using the TeloTAGGG polymerase chain reaction (PCR) ELISAPLUS kit (Roche, Mannheim, Germany) according to the manufacturer's protocol. For each sample, 0.5 μg of total protein was added to the PCR reaction. Relating the enzyme-linked immunosorbent assay (ELISA) signal of the sample to the one obtained by a control template with a known number of telomeric repeats, relative telomerase activity values were calculated and translated in fold of telomerase activity of the 293T cell line–based phoenix ampho cells carried along with each test. RNase-treated cellular lysates were used as negative controls for each sample. In addition, 20 μl of PCR product was resolved by a 12% polyacrylamide gel and visualized on a Biotin Luminescent Detection Kit (Roche) according to the manufacturer's protocol.
Flow-Fluorescence In Situ Hybridization (Flow-FISH)
To determine average telomere length in individual cells, cells were hybridized in situ with a fluorescent telomere-specific peptide nucleic acid probe, as has been described recently . Telomere length was expressed in telomere fluorescence units based on calibration experiments using TRF (telomere restriction fragment) length analysis by Southern blotting.
Hematopoietic progenitors were assayed by plating aliquots of cells in methylcellulose medium (GF H4434, Stem Cell Technologies) supplemented with 50 ng/ml SCF, 10 ng/ml GM-CSF, 10 ng/ml IL-3, and 3 U/ml erythropoietin (Epo). After 2 weeks of incubation at 37°C, burst-forming units erythroid (BFU-E), colony-forming units granulocyte-macrophage (CFU-GM), and CFU–granulocyte-erythroid-macrophage-megacaryocyte (CFU-GEMM) were counted.
Data analysis was performed using Microsoft Excel and Microcal Origin software. Results are shown as means ± standard error of values obtained in independent experiments. Wilcoxon rank-sum test was used to determine statistical significance.
Ectopic Transfer of hTERT in CD34+ CB Cells Increases the Levels of Telomerase Activity
To evaluate the function of two hTERTs containing retroviral vectors, we transduced primary human fibroblast cells that are telomerase-negative with the pBABE-hTERT-IRES-GFP and the pOS-hTERT-IRES-GFP constructs, respectively. Similar to previous reports [26, 27], we found for both vectors that ectopic transfer of hTERT-induced telomerase activity, which was accompanied by substantial telomere elongation in the range of 2–2.5 kb as measured by flow-fluorescence in situ hybridization (flow-FISH; Figs. 1A and 1B). As a result, both constructs were able to immortalize human diploid fibroblasts, whereas control cells entered senescence after approximately 60 population doublings (Fig. 1C).
Next, CD34+ CB cells were transduced with the retroviral vector pBABE-hTERT-IRES-GFP. GFP+ cells were fluorescence-activated cell (FAC)–sorted and cultured in serum-free medium supplemented with SCF, Flt-3, IL-3, IL-6, and G-CSF. Transgene expression was stable, as indicated by GFP expression above 90% at the end of the culture (data not shown).
The endogenous telomerase activity of enriched CD34+ cells was moderate, being 51 ± 9% (n = 5) of the activity found in a representative telomerase-positive immortal cell line (ϕNX). During the subsequent 6 days of cell expansion, enzyme activity was strikingly increased by more than 15.3 ± 6.5-fold (n = 5) compared with day 0, which was followed by a continuous decline thereafter. In CD34+ cells transduced with the hTERT gene, telomerase activity was significantly elevated relative to control cells throughout the culture (Figs. 2A and 2C). Considering the decline of endogenous telomerase activity, a peak fold increase of telomerase activity in ectopically hTERT-expressing cells was achieved by day 22, reaching 10.4 ± 3.1-fold (p < .05, n = 7) relative to vector-only transduced cells (Fig. 2B). Thus, the remaining ectopic telomerase activity was comparable with the levels that are found in immortalized cells, such as the phoenix cell line. Similar results were obtained using the pOS-hTERT-IRES-GFP vector (n = 3; data not shown).
Telomere Length Dynamics in hTERT-Transduced CD34+ CB Cells
The average length of telomeric repeats was measured in transduced and nontransduced CD34+ cells at different time points during in vitro expansion using the flow-FISH method. Figure 3A shows typical telomere fluorescence histograms of hTERT overexpressing and control cells in the course of the expansion culture. The mean telomere length of enriched CD34+ CB cells was found to be 9.2 ± 0.2 kb (n = 6; Fig. 3B). Upon cell expansion, the mean telomere length of control cells was decreased by 1.1 ± 0.4 kb between days 10 and 22 of the culture (p < .05, n = 6). In CB cells ectopically expressing hTERT, the mean telomere length was found to be slightly increased in the range of 200 to 600 bp during different time points of cell expansion, which was significant by days 10 and 14 (p < .05) but not by day 22 (p = .3). However, despite increased telomerase levels, telomere attrition was not prevented in hTERT-transduced CB cells, resulting in a loss of 1.3 ± 0.4 kb over the culture period (p < .05, n = 6; Fig. 3B). Thus, overexpression of telomerase does not seem to stabilize the mean telomere length in dividing hematopoietic progenitor cells.
Modulation of Telomerase Activity Does Not Influence Telomere Length Dynamics inAC133+ CB Cells
To exclude that the heterogeneity within the selected CD34+ cell population affects the outcome of observed telomere length dynamics, an additional set of experiments (n = 4) was performed in which CD34+ CB cells were sorted for GFP and AC133-PE fluorescence after transduction with the pOS-hTERT-IRES-GFP vector. In addition, transduction was performed with the same construct containing a dominant-negative mutant of hTERT (DN-hTERT).
As expected, AC133+GFP+ cells, which were transduced with the vector containing the wild-type hTERT gene, showed elevated levels of telomerase activity, whereas those transduced with the dominant-negative mutant had strikingly reduced or absent telomerase activity (Fig. 4A). Interestingly, the mean telomere length was not significantly affected (Fig. 4B) despite these highly different levels of telomerase activity, which confirms our observation that modulation of telomerase activity itself has only limited impact on telomere length dynamics in hematopoietic progenitor cells during in vitro culture.
Reduced Colony Formation in CB Cells with Loss of Telomerase Activity
To evaluate the formation of CFU in culture (CFU-C), FAC-sortedAC133+GFP CB cells were plated in methylcellulose supplemented with SCF, GM-CSF, IL-3, and Epo (Fig. 5A). We observed reduced colony formation in cells expressing the dominant-negative mutant form of hTERT in each of the four CB samples. Colony-forming cell (CFC) activity was also assessed for two CB samples after 7 days in expansion culture. Again, DN-hTERT overexpression was accompanied by reduced CFU-C (Fig. 5B). Interestingly, overexpression of hTERT did not increase the colony numbers for CFU-GM, CFU-GEMM, and total CFC compared with control cells with endogenous telomerase activity. However, we found a trend in increasing numbers of BFU-E at the two investigated time points. Thus, the absence of telomerase activity may impair colony formation in general, whereas additional levels beyond endogenous activity seem not to increase the clonogenic capacity, with the exception of the BFU-E lineage.
Telomere manipulation has become an interesting topic with regard to attempts to increase the proliferative lifespan of normal human cells. Ectopic expression of hTERT has been shown to maintain or elongate telomeric repeats in some cell types, leading to an immortal phenotype . In this study, we demonstrate that augmentation of telomerase activity in human hematopoietic progenitor cells from CB had only minimal impact on telomere-length dynamics and did not prevent overall telomere shortening during in vitro expansion. This finding is in contrast to reports on different cell types, such as retinal pigment cells , T lymphocytes , and bone marrow stromal cells , in which telomeres are extended up to several kilobases when overexpressing hTERT. Because our vector constructs were proven to be functional in primary fibroblast cells by extending the mean telomere length by more than 2 kb, it is very unlikely that the observed effects are due to technical limitations.
Several reasons might be responsible for the failure to modulate telomere length in hematopoietic progenitor cells using a genetic approach with hTERT transfer. In general, telomere length homeostasis is known to result from a balance of lengthening and shortening factors . Although the telomerase catalytic component produces a strong lengthening activity, other factors, including telomere binding proteins such as Pin2/TRF1 , TRF2 , and TIN2 , are involved in establishing a telomere-length equilibrium. Most likely, the telomere-length homeostasis is more tightly regulated in hematopoietic progenitor cells compared with other cell types that are telomerase negative. Another explanation for our observations could be related to the fact that despite high telomerase activity found in the telomerase repeat amplification protocol assay, the accessibility of the enzyme at the telomeric end could be reduced, which might be attributable to factors such as the end-binding protein POT1  or hEST1A . So far, the role of these factors in hematopoietic cells has not yet been determined. In addition, it has to be taken into account that telomere length in CB cells is quite high (9 to 10 kb). Therefore, a strong negative-feedback mechanism mediated by TRF1 might limit the action of telomerase to keep telomere length in a narrow range .
The role of telomerase in hematopoietic stem cells is incompletely understood. In mice, it has been suggested that it serves to slow telomere shortening and confers extended proliferative capacity for a highly proliferative tissue . In humans, even partial telomerase deficiency has been associated with early onset of bone marrow failure, such as for patients with the disease dyskeratosis congenita [38, 39]. In our experimental system, ablation of telomerase activity was associated with general reduction of CFU-C numbers, which is compatible with data on telomerase-deficient HSCs in mice . Because CB cells have rather long telomeres that should not signal telomere dysfunction, it is tempting to speculate that telomerase may play a direct role in cell-cycle progression of hematopoietic progenitor cells. Interestingly, augmentation of telomerase activity beyond the endogenous levels did not lead to a general increase in proliferative capacity, suggesting that only a minimum of activity might be required for that function. Similarly, it has been recently demonstrated in a transgenic mouse model system that over expression of hTERT did not extend the transplantation capacity of HSCs, which suggests that telomere-independent barriers may limit the transplantation capacity of murine stem cells . Interestingly, telomere length was kept stable in murine HSCs, which had approximately fourfold increased telomerase activity, which is in contrast to the failure of telomere maintenance in our human experimental system despite 10-fold increase of telomerase activity. Fundamental differences in telomere biology between mouse and human may possibly explain this differential outcome in telomere length dynamics between both species [41, 42].
Finally, our data provide evidence that over expression of hTERT might selectively increase the rate of erythroid colonies, which suggests that the enzyme might have an impact on lineage differentiation. Interestingly, evidence is now accumulating that additional functions of telomerase exist beyond a simple telomere-length maintenance activity [43–45]. Nevertheless, more data are necessary to confirm this theory with respect to the differentiation of hematopoietic progenitor cells, because we finally cannot exclude the possibility that ectopic hTERT is more expressed in differentiated cells than by expression in progenitor cells.
We show that modulation of telomerase activity alone does not have an impact on telomere dynamics in hematopoietic progenitor cells, which is in contrast to various other cell types. Therefore, strategies that aim at telomere lengthening to rejuvenate adult stem cells require additional exploration of mechanisms that are involved in telomere length regulation. Ideally, such studies should be performed in context to telomere maintenance mechanisms in embryonic stem cells. Furthermore, our data reveal that telomerase activity might play a direct role in proliferation and differentiation of hematopoietic progenitor cells. More studies are necessary to define the role of other factors such as telomere binding proteins with regard to this process in the hematopoietic system.
We gratefully acknowledge the excellent technical assistance of I. Skatulla. We also thank Dr. A. Dwenger for providing blood samples, Dr. M. Engelhardt and Dr. M. Zijlmans for helpful discussions and critical comments on the manuscript, and Professor Mertelsmann for his support. In addition, we are indebted to Dr. R. Weinberg and Dr. H. Vaziri for providing the retroviral hTERT construct. This work was supported by grants from the Deutsche Forschungsgemeinschaft (SFB 364) and the European Union (QLG1-CT-1999-01341).