Search

We report for the first time potential pathogenic mutations in a known telomeric DNA binding protein factor (TIN2) in patients with acquired aplastic anemia. We examined samples collected from 142 AA patients and 289 healthy controls and found several natural sequence changes in the TINF2 gene, which encodes for the TIN2 protein. Lymphocytes collected from patients with TINF2 variant alleles appeared to grow slower in culture than cells collected from either healthy individuals or those who carried inconsequential polymorphic TINF2 sequence changes (SNPs), consistent with the fact that telomere lengths in these patients were shorter than those of healthy individuals of similar age. We introduced natural mutations in the TIN2 promoter region into a TIN2-promoter driven luciferase reporter plasmid and examined their effect on driving luciferase reporter expression. Using a similar approach, we defined both the regulatory DNA elements and transcriptional factors that control TINF2 gene expression under normal physiological condition. Results obtained from these studies were validated in gel-shift (EMSA) and chromatin immunoprecipitation (ChIP) assays. In addition to mutations in the promoter region, numerous natural mutations were identified in the coding region of the TINF2 gene in patients with AA or Dyskeratosis Congenita (DC). We used co-immunoprecipitation assays to test the effect of natural mutations in disrupting TIN2's interaction with its known protein partners TRF1, TRF2 and TPP1 of the shelterin complex that protects chromosome ends. The results obtained from these studies will be discussed. To our knowledge, this is the first comprehensive functional analysis of all known mutations in the TINF2 gene in patients with AA or DC. This study offers an insight into how TINF2 mutations can lead to telomere shortening effect and to limited marrow stem-cell reserve and renewal capacity in patients with various forms of bone-marrow failure syndromes.

Dyskeratosis Congenita (DC) is a marrow failure syndrome characterized by skin and nail abnormalities, oral leukoplakia and very short telomeres in circulating leukocytes. Heritable defects in telomere maintenance have been directly implicated in DC by the discovery of mutations in genes encoding components of the telomerase complex: DKC1, TERT, and TERCas well as mutations in the gene encoding the telomere binding protein TINF2. Defective telomeres in DC result in impaired hematopoiesis and predispose to myeloproliferative disorders. Heritable mutations in TERTand TERChave also been implicated in patients presenting with aplastic anemia (AA) and idiopathic pulmonary fibrosis (IPF) without clinical signs of DC. Because short telomeres appear to be associated with increased risks for various human cancers, including head and neck cancer, we sequenced TERTand TERCin two patients with oral carcinoma and anemia.

Heritable mutations in the genes that encode the three minimal components of the human telomerase complex, hTERT, hTERC and DKC are known to give rise to Dyskeratosis Congenita (DC), a rare disorder characterized by skin pigmentation abnormalities, nail dystrophy and leukoplakia. Telomeres in leukocytes of patients with DC are invariably very short and patients typically succumb to consequences of bone marrow failure, pulmonary fibrosis or malignancies. Other genetic defects including mutations in the telomeric protein TINF2 are also known to give rise to DC. However, not all individuals with mutations in “telomere maintenance” genes such as TERT, TERC, DKC and TINF2 will develop clinical symptoms during their lifetime and some patients, without clinical signs of DC, present with aplastic anemia (AA) and idiopathic pulmonary fibrosis (IPF). It was previously shown that hypomorphic mutations in hTERT are 3-fold more common in patients with acute myeloid leukemia (AML) than in controls (Calado et al., ASH abstracts 2007 110: 16). Together with the increased incidence of malignancies in DC these observations suggest that telomere dysfunction can trigger dysplastic as well as neoplastic disorders, most likely because progressive telomere loss results in loss of normal cells and thereby selects for cells with defects in the DNA damage checkpoint(s) that are normally triggered when chromosome ends have insufficient telomere repeats. Such cells are expected to have DNA repair defects and their malignant evolution could be facilitated further by telomere dysfunction triggering cycles of chromosome fusions/bridge/breakage before telomerase is eventually upregulated. In view of these considerations and the important role of telomeres in B cell biology (with telomeres being elongated in the germinal centre and memory B cells having longer telomeres than naïve B cells) we postulated that heritable genetic defects in telomere maintenance could predispose to B cell malignancies as well as AML. To test this hypothesis we sequenced TERC and TERT genes in 80 consecutive CLL patients. No mutations in TERC were found. Sequence variants in TERT were identified in 14 patients with one patient, a compound heterozygous, carrying 2 separate mutations. 5 of the 80 CLL patients carried the A279T TERT variant but this allele was also present in ~ 3% of control individuals. This TERT allele did not significantly reduce telomerase activity in telomerase reconstitution experiments measured by TRAP assay. All other TERT sequence variants that we found in CLL appear to be hypomorhic mutations (that reduce but not completely disable telomerase reverse transcriptase activity) and all were previously described in DC, AA, AML and IPF. Two common variants, D441E and A1062T were screened by high throughput dotblotting of DNA from CLL patients and frequencies of 3/142 (2.1 %) and 12/195 (6.2 %) were found respectively. The presumed germline origin of the TERT mutations in CLL needs to be confirmed. Our results indicate that hypomorphic TERT mutations are common in CLL and contribute to disease in over 10% of patients.

Telomeres protect the ends of chromosomes, shorten with age, and are very short in dyskeratosis congenita (DC), an inherited bone marrow failure syndrome (IBMFS) associated with mutations in telomere biology genes. “Short telomeres” were reported in Fanconi Anemia (FA), Diamond-Blackfan Anemia (DBA) and Shwachman-Diamond Syndrome (SDS) using telomere restriction fragment length or Q-FISH assays of total leukocyte or mononuclear cell DNA. These reports focused on group averages, not results from individual patients. Our objective was to determine which categories of IBMFS patients have very short telomeres, and in which leukocyte subsets, using a more sensitive and specific assay. Telomere length was measured in granulocytes, lymphocytes, naïve T-cells, memory T-cells, B-cells, and NK cells using automated multicolor flow fluorescence in situ hybridization (FISH). We previously showed that very short telomeres (<1st percentile for age) in lymphocytes, naïve T-cells, and B-cells were sensitive and specific for the diagnosis of DC (Alter et al, Blood110:1439, 2007). Granulocytes were not specific, since in that study neutropenic patients without DC often had very short granulocyte telomeres. The current study included 53 DC patients and 87 relatives, 19 non-transplanted FA patients and 12 relatives, 21 DBA patients and 11 relatives, and 7 SDS patients and 7 relatives. There were 400 normal controls, ages 0 to 100; very short telomeres were defined as <1st percentile of normal for age. Z-scores were calculated to adjust for age; 0 is normal and -2 or below is significantly short (i.e. <2 standard deviations below the mean). Lymphocyte telomeres were <1st percentile (very short) in 96% of DC patients, 7% of DC relatives, 16% of FA patients, 5% of DBA patients, 14% of SDS patients, and none of the non-DC relatives. Granulocyte telomeres were very short in 94% of DC patients, 16% of DC relatives, 44% of FA patients, 19% of DBA patients, 14% of SDS patients, 8% of FA relatives, and none of the non-DC relatives. All three hallmark leukocyte subsets (lymphocytes, naïve T-cells, and B-cells) were very short in 81% of DC patients, 5% of DC relatives, 5% of DBA patients, and none of the FA or SDS patients or their relatives. Although 4 out of 47 (9%) IBMFS other than DC had very short telomeres in lymphocytes, only one, a patient with DBA (without mutations in known DC genes) had very short telomeres in all 3 of the hallmark lineages. Four of 87 DC relatives also had very short telomeres in all 3 lineages; they may be silent carriers in families in which DC genes have not yet been identified. The mean Z-score for lymphocyte telomeres in DC was -4.7; it was above -2 in all other categories (-0.3 in FA, -0.9 in DBA, -1 in SDS). The mean Z-score for granulocytes was -4.2 in DC, -1.9 in FA, -1.2 in DBA, and -1.5 in SDS. Thus, based on the Z-scores, the average telomere length was very short in DC lymphocytes and granulocytes; normal in FA lymphocytes and borderline in FA granulocytes; and normal in DBA and SDS lymphocytes and granulocytes. Flow-FISH provides analysis of multiple cell types in individual patients, and identification of specific individuals with very short telomeres, as well as the age-adjusted mean telomere length in a diagnostic subset. No FA or SDS patients, and only one with DBA, met the DC diagnostic criteria of very short telomeres in three or more lymphocyte subsets. The telomere length deficit in DC appears to be more severe and more frequent than in the other inherited bone marrow failure syndromes.

Telomeres protect the ends of chromosomes, shorten with age, and are very short in dyskeratosis congenita (DC), an inherited bone marrow failure syndrome (IBMFS) associated with mutations in telomere biology genes. “Short telomeres” were reported in Fanconi Anemia (FA), Diamond-Blackfan Anemia (DBA) and Shwachman-Diamond Syndrome (SDS) using telomere restriction fragment length or Q-FISH assays of total leukocyte or mononuclear cell DNA. These reports focused on group averages, not results from individual patients. Our objective was to determine which categories of IBMFS patients have very short telomeres, and in which leukocyte subsets, using a more sensitive and specific assay. Telomere length was measured in granulocytes, lymphocytes, naïve T-cells, memory T-cells, B-cells, and NK cells using automated multicolor flow fluorescence in situhybridization (FISH). We previously showed that very short telomeres (<1stpercentile for age) in lymphocytes, naïve T-cells, and B-cells were sensitive and specific for the diagnosis of DC (Alter et al, Blood 110:1439, 2007). Granulocytes were not specific, since in that study neutropenic patients without DC often had very short granulocyte telomeres. The current study included 53 DC patients and 87 relatives, 19 non-transplanted FA patients and 12 relatives, 21 DBA patients and 11 relatives, and 7 SDS patients and 7 relatives. There were 400 normal controls, ages 0 to 100; very shorttelomeres were defined as <1stpercentile of normal for age. Z-scores were calculated to adjust for age; 0 is normal and -2 or below is significantly short (i.e. <2 standard deviations below the mean). Lymphocyte telomeres were <1stpercentile (very short) in 96% of DC patients, 7% of DC relatives, 16% of FA patients, 5% of DBA patients, 14% of SDS patients, and none of the non-DC relatives. Granulocyte telomeres were very short in 94% of DC patients, 16% of DC relatives, 44% of FA patients, 19% of DBA patients, 14% of SDS patients, 8% of FA relatives, and none of the non-DC relatives. All three hallmark leukocyte subsets (lymphocytes, naïve T-cells, and B-cells) were very short in 81% of DC patients, 5% of DC relatives, 5% of DBA patients, and none of the FA or SDS patients or their relatives. Although 4 out of 47 (9%) IBMFS other than DC had very short telomeres in lymphocytes, only one, a patient with DBA (without mutations in known DC genes) had very short telomeres in all 3 of the hallmark lineages. Four of 87 DC relatives also had very short telomeres in all 3 lineages; they may be silent carriers in families in which DC genes have not yet been identified. The mean Z-score for lymphocyte telomeres in DC was -4.7; it was above -2 in all other categories (-0.3 in FA, -0.9 in DBA, -1 in SDS). The mean Z-score for granulocytes was -4.2 in DC, -1.9 in FA, -1.2 in DBA, and -1.5 in SDS. Thus, based on the Z-scores, the average telomere length was very short in DC lymphocytes and granulocytes; normal in FA lymphocytes and borderline in FA granulocytes; and normal in DBA and SDS lymphocytes and granulocytes. Flow-FISH provides analysis of multiple cell types in individual patients, and identification of specific individuals with very short telomeres, as well as the age-adjusted mean telomere length in a diagnostic subset. NoFA or SDS patients, and only one with DBA, met the DC diagnostic criteria of very short telomeres in three or more lymphocyte subsets. The telomere length deficit in DC appears to be more severe and more frequent than in the other inherited bone marrow failure syndromes.

Dyskeratosis congenita (DC) is an inherited bone marrow failure syndrome characterized by the triad of abnormal nails, lacey reticular skin pigmentation, and oral leukoplakia. Patients with DC are at high risk of developing aplastic anemia, myelodysplastic syndrome and leukemia. Diagnosis of DC is challenging due to variability of the triad and heterogeneous clinical findings such as pulmonary and liver disease, avascular necrosis, esophageal or urethral stenosis and development delay. The unifying feature in DC is exceedingly short telomere lengths and defects in telomere biology. Gene mutations have been identified in DKC1 (X-linked), TERC and TERT (dominant, AD) and NOP10 (recessive), but approximately 60% of DC patients lack a known mutation. We identified a non-consanguineous family with AD DC and no mutations in DKCI, TERC or TERT. The first DC cases were monozygotic twin brothers (now deceased). Telomere length was determined by flow-FISH on the twins’ children (6 affected, 4 unaffected), wives, their 8 siblings and parents. Despite variable clinical phenotypes, the 6 affected individuals (1 female, 5 male) all had very short telomere lengths (<1st%ile for age). Unaffected relatives had normal telomere lengths. A single nucleotide polymorphism (SNP) genome-wide linkage screen (Human Linkage IVb, Illumina, Inc) was conducted using telomere length <1st%ile as the affected phenotype. SNPLINK was used to remove SNPs in linkage disequilibrium (D′=0.7, R2=0.4). Data were analyzed with GeneHunter under a parametric, AD, rare, highly-penetrant disease model. Evidence favoring linkage was found in a 17 megabase (Mb) region on chromosome 2p and a 3.1 Mb region on chromosome14q (LOD score=2.62 at both sites). Bi-directional sequence analysis of the two best candidate genes, DDX1 (2p) and TINF2 (14q) was conducted to identify mutations. A novel mutation, K280E, in TINF2 (protein name TIN2) was identified in the 6 living, affected family members but not in the 8 unaffected relatives, suggesting inheritance from the affected fathers. There were no mutations in DDX1. TINF2 was sequenced in 8 additional, unrelated DC probands without DKCI, TERC or TERT mutations and 7 with known mutations. An R282H mutation was present in 3 unrelated DC probands (1 Hoyeraal-Hreidarsson, 1 Revesz Syndrome). Another DC patient had an R282S mutation. TINF2 mutations were not present in unaffected relatives, DC probands with mutations in DKC1, TERC or TERT, or 298 controls. The mutation prevalence in DC probands represented in our cohort of patients with DC are DKC1 (18.8%), TERC (18.8%), TERT (6.2%), and TINF2 (31.2%). As a component of shelterin, the protein complex that stabilizes telomeres, TIN2 is highly evolutionarily conserved. It serves as a bridge between the three primary telomere DNA-binding proteins, TRF1, TRF2 and POT1 (via TPP1). In silico analyses predict that K280E, R282H and R282S are deleterious mutations. Functional studies are underway to characterize possible effects of these mutations on shelterin protein interactions. By focusing on telomere length as the affected phenotype, instead of the heterogeneous clinical features present in DC patients, we identified mutations in TINF2 and further validated telomere length as a diagnostic test for DC. This study demonstrates that TINF2 is the 5th gene mutated in DC, the 1st in Revesz Syndrome, and the 1st shelterin complex gene mutated in human disease.

Telomeres protect chromosome ends from end-to-end fusion and recombination but are gradually eroded with proliferation. Critically short telomeres are repaired by the telomerase complex and mutations in telomerase complex genes (TERT and TERC) are associated with dyskeratosis congenita and acquired aplastic anemia. Telomere shortening eventually results in cell proliferation arrest, apoptosis, or genomic instability, and patients with dyskeratosis congenita and acquired aplastic anemia are at risk for developing leukemia and other malignancies. We therefore investigated whether TERT and TERC gene mutations were associated with acute myeloid leukemia (AML) by screening 100 consecutive Brazilian patients diagnosed with AML (excluding acute promyelocytic leukemia) at the same institution and 198 ethnic-, age-, and sex-matched healthy volunteers. Eight patients carried a non-synonymous TERT gene variant; one was homozygous and five heterozygous for A1062T, one was heterozygous for H412Y, and another was heterozygous for a novel R522K TERT mutation. None of these gene variants was present in matched controls (Fisher’s exact test, P=0.0001). We validated our results by screening an additional 89 AML patients from the MD Anderson Cancer Center, selected based on cytogenetic status and 528 healthy controls. Four of these 89 patients had a non-synonymous TERT gene variant (one homozygous, three heterozygous), an incidence higher than in controls (P=0.028). To further address the higher prevalence of the A1062T gene variant in AML, we screened a total of 1,111 controls, and found a 3.8 times higher A1062T allele frequency in AML than in controls (P=0.002). The germ-line origin of mutations was established by mutation detection in non-hematopoietic tissues and in relatives. Cytogenetics were available for four TERT-mutant patients in the Brazilian cohort: two had inv(16), one had t(5;11)(q35;q13) and del(10)(p15), and another had complex karyotype (including trisomy 8). In the MD Anderson cohort, 2 patients had inv(16) and two had trisomy 8. The incidence of TERT mutations was higher in patients with trisomy 8 (2/10) or inv (16) (2/22) than in other patients (0/57; P =0.02 for trisomy 8 and 0.08 for inv(16) vs. other). Telomere lengths of blast cells were extremely short (median length, 3.1 kb; range, 2.4-5.9 kb). Vectors containing TERT mutants were transfected into VA13 cells and telomerase activity of transfected cell lysates, measured by the fluorescent telomere repeat amplification protocol (TRAP) assay, demonstrated that AML-associated TERT gene variants resulted in significantly reduced telomerase function in comparison to wild-type TERT by haploinsufficiency. Our results indicate that constitutional TERT gene mutations are risk factors for AML. Abnormal telomerase function of hematopoietic stem cells may result in short and dysfunctional telomeres, facilitating genomic instability, aneuploidy, and probably contributing to an early stage of leukemogenesis.

Dyskeratosis congenita (DC) is an inherited multisystem disorder of premature aging, typically characterized by bone marrow failure, mucosal leukoplakia, abnormal skin pigmentation, and nail dystrophy. The X-linked and autosomal dominant forms of DC are associated with mutations in genes that affect telomerase activity resulting in a decrease in telomere length. DC, like other bone marrow failure disorders, is associated with ineffective hematopoiesis and a cancer predisposition. Standard treatment of bone marrow failure or cancer requires cytotoxic therapy, and clinical observations suggest DC patients have an increased sensitivity to cytotoxic therapy. To explain this, we hypothesized that the short telomeres in somatic cells from DC patients could alter the activity and/or expression of several proteins involved in DNA repair or the response to cellular stress including p16, p53 and p21. Lymphocytes from five DC subjects and age-matched controls were stimulated to grow in vitro in the presence of various cytotoxic agents with different modes of action, including Taxol (antimitotic agent and microtubule inhibitor) and Etoposide (topoisomerase inhibitor and DNA damaging agent). In addition, we tested fibroblasts and keratinocyte extracted from skin biopsies from DC and control subjects that were serially passaged. Cellular proliferation and cell death were monitored by cell counts and flow cytometry. Western blotting was used to measure steady state and DNA damage- induced expression of tumor suppressor protein p53 and other proteins involved in DNA damage response signaling pathway, including p16 and p21 in relation to telomere length. Results of flow cytometry accompanied by direct visualization showed a decreased proliferation of DC lymphocytes compared to normal cells, and this growth disadvantage was further accentuated following cell exposure to cytotoxic agents. DC lymphocytes exposed to 10−6 M Taxol showed a decrease in cellular proliferation between 3 and 8 fold while normal control cells exposed to the same agents exhibited only a 3 to 4 fold decrease in cell growth. Similarly DC lymphocytes exposed to Etoposide were inhibited to a greater extent than control cells. Western blot analysis of whole cell lysates indicated a difference in DNA damage response proteins. Of note, lymphocytes from several DC subjects exposed to Taxol did not upregulate p53 expression, while inducible levels were noted in Taxol-treated control cells. In contrast, DC and control lymphocytes exposed to Etoposide upregulated p53 in a similar dose dependent manner. No differences were noted in DC versus control lymphocytes with regards to basal or chemotherapy induced p16 expression. Interestingly, late passage DC fibroblasts displayed enhanced basal expression of p16. These results support the clinical observation of increased “chemosensitivity” in DC subjects and suggest that diminished telomerase activity and premature telomere shortening may interfere with normal DNA damage and stress response pathways. These data are also consistent with our finding that DC fibroblasts, keratinocytes, and lymphocytes have a reduced cell proliferative lifespan. Further studies are needed to dissect the role of telomeres in the cellular response to various types of DNA damage.

Dyskeratosis Congenita (DC) is a bone marrow (BM) failure syndrome characterized by progressive telomere shortening. The discovery of mutations in genes encoding components of the telomerase complex in DC has implicated defective telomere maintenance in the disease pathogenesis. However, it is not known if defective telomeres in DC result in a qualitative impairment of hematopoiesis and/or a decrease in the maintenance of very primitive hematopoietic cells. To address this question, we characterized CD34+ cells present in BM aspirates and granulocyte colony stimulating factor (G-CSF) -mobilized peripheral blood (mPB) cells harvested from 5 DC patients prior to the onset of severe BM failure. These 5 patients were from 3 successive generations of a single family and all had an autosomal dominant mutation in the telomerase gene (TERC). The BM was found to be hypocellular in all patients, with a consistently reduced percentage of CD34+ cells - lowest in the two 3rd generation subjects (0.05% and 0.16% vs 0.5%–1% in normal controls). The target for G-CSF-mobilized cell harvests (2x106 CD34+ cells/kg) was met for 2 of the 5 DC patients, with only 5x105 CD34+ cells/kg obtained from the 2 DC patients with a longstanding history of thrombocytopenia. However, within the CD34+ population, the proportions of cells co-expressing various lineage markers were normal and the frequencies of granulopoietic, erythroid and multi-lineage colony-forming cells (CFCs) were similar to those of normal donors, suggesting that the commitment and differentiation programs in these DC patients were not grossly perturbed. However, the fact that their BMs were hypocellular with a further reduced content of CD34+ cells indicates that total CFC numbers were markedly decreased. The total number of more primitive cells detectable as long term culture-initiating cells (LTC-ICs) were decreased a further order of magnitude as indicated by a 10-fold reduction in their frequency within the CD34+ compartment by comparison to normal mPB CD34+ cells. Telomere length measurements were performed on cells from day 4 cultures of CD34+ mPB DC cells using quantitative FISH (Q-FISH) and single telomere end length analysis (STELA). Telomere length in a 2nd generation patient was reduced (7.2 Kb) compared to an age-matched control (12.4 Kb) and this difference was accentuated in two 3rd generation patients (4.6 and 3.4 Kb) in line with progressive telomere loss with successive generations. Telomeres with <1.5 kb repeats were readily observed by STELA and were increased in DC subjects (up to 9.5%). These findings provide the first description of telomere length and functional analyses of primitive hematopoietic cells from DC patients. The data support the concept that BM failure in DC subjects results from a loss of very primitive hematopoietic cells triggered by defective telomeres rather than a qualitative impairment in the ability of hematopoietic cells to differentiate. Assays of primitive cells and telomere length may thus be useful parameters to predict both the time of onset and the degree of BM failure in DC patients.

Mutations in telomere repair complex genes TERT (encoding telomerase reverse transcriptase) and TERC (telomerase RNA component) are associated with bone marrow failure, especially acquired aplastic anemia and dyskeratosis congenita. Low telomerase activity leads to short telomeres of leukocytes, predisposing highly proliferative tissues such as the bone marrow to early senescence and exhaustion of the stem cell compartment. Telomere repair gene mutations have been suggested to result in disease anticipation, defined as earlier and/or worsening clinical manifestations in successive generations. We have identified a six-generation pedigree in a large Mennonite family carrying a novel TERT mutation (K570N), which localizes in the catalytic domain with reverse transcriptase activity (RT domain). The index patient is a 26 year old male dairy farmer with a ten-year history of severe aplastic anemia (5% bone marrow cellularity) unresponsive to immunosuppression. A long history of hematologic diseases was well known and named in the family--the patient’s paternal great-great-grandmother had died of a severe blood disorder at age 65 years. However, the great-grandmother and the grandfather had never presented any hematological disease. The patient’s father had myelodysplastic syndrome at age 33 years, evolving to acute myeloid leukemia and death due to failure to recover blood counts after chemotherapy. One of the proband’s paternal aunts had aplastic anemia develop when she was a young woman and has been transfusion-independent for decades in response to chronic androgen therapy. A second proband’s paternal aunt underwent a liver transplant at age 20 for submassive hepatic necrosis with fibrosis. A third proband’s paternal aunt has macrocytosis only at age 47. Two sisters (ages 21 and 23) also have macrocytosis in the absence of other hematological abnormality, and two other sisters are healthy. Genetic analysis showed that TERT K570N mutation is present in the patient’s paternal (including grandfather, three aunts and two affected sisters) but not maternal relatives (making his father an obligatory carrier). The patient’s oldest of three sons, now age four years, has the mutation but is asymptomatic. There was no nail dystrophy, leukoplakia or skin hyperpigmentation in any of the TERT K570N carriers; although the index patient and some of his relatives showed early graying of hair, this characteristic did not track with the mutation. Telomere shortening of leukocytes, as measured by Flow-FISH, tracked to the mutation in three generations analyzed, being shortest in the proband and in his aunt with marrow failure. Mutagenized TERT vectors transfected into telomerase-deficient VA13 cell lines yielded no telomerase activity using the telomeric repeat amplification protocol (TRAP) assay, whereas when wild-type TERT vectors were co-transfected, telomerase activity was approximately half of wild-type transfected only, indicating haploinsufficiency as a mechanism of telomere shortening. Our results confirm the association between aplastic anemia and TERT mutations. The pattern of hematologic disease in this kindred does not support disease anticipation in TERT mutations. Most remarkably, there is a likely relationship between a telomerase gene mutation and hematological malignancy and severe liver disease.

T-cell prolymphocytic leukemia (T-PLL) is a rare aggressive lymphoproliferative disease characterized by the expansion of a T-cell clone derived from immuno-competent post-thymic T-lymphocytes. Important mechanisms involved in expansion of human malignant cells are reactivation of telomerase, an enzyme complex, which is able to compensate the loss of telomere repeats by cell division, and maintenance or elongation of telomere length. To investigate the role of telomeres and telomerase we measured telomere length by automated multicolor flow-FISH and telomerase activity by telomeric repeat amplification protocol in subsets of peripheral blood leukocytes from 11 newly diagnosed or relapsed patients with sporadic T-PLL. In addition, we analyzed the effect of the selective telomerase inhibitor BIBR1532 on T-PLL cells in short-term culture assays. The average telomere length in the clonal T-cells of all samples analyzed was extremely short (mean ± std: 1.53 kb ± 0.65 kb) compared to the non-clonal T-cells (5.03 kb ± 0.71 kb; p=0.012). The average telomere length for B-cells in these patients was 6.37 kb ± 0.71 kb (n=6). Telomere length values of the clonal T-cells were all below the 1st percentile of telomere length values observed in T-cells from healthy aged-matched controls whereas non-clonal T-cells and B-cells fell between the 1st and 99th percentile of the normal distribution. Interestingly, telomere length in the clonal T-cells remained stably short at 1.0 kb ± 0.6 kb without further telomere loss in one patient over a period of 18 months. No cell doublets indicative of fused or bridged chromosomes and telomere dysfunction were observed. Clonal T-cells exhibited high levels of telomerase activity almost comparable to levels of the positive control K562 whereas there was no measurable telomerase activity in normal, unstimulated T-cells. Telomerase levels even correlated inversely with telomere length in clonal T-cells (r=-0.91, n=6). In addition, we could induce a dose-dependent cytotoxicity of T-PLL cells with the telomerase inhibitor BIBR1532 (viable cells as percentage of untreated controls (viability index) in % after 10 days of cell culture (mean ± std) with 0, 10, 40, 100 µM BIBR1532: 100 ± 0, 54 ± 14.3, 27.5 ± 11.7, 2.6 ± 1.6, n=6) whereas no effect was observed in normal, unstimulated T-cells (viability index in % (mean ± std) after 10 days with 0, 10, 40, 100 µM BIBR1532: 100 ± 0, 100.8 ± 8.6, 96.9 ± 7.9, 103.1 ± 22.4, n=3). In summary, clonal T-cells in T-PLL exhibit extremely short telomeres which could explain the genomic instability with cytogenetic aberrations. The high levels of telomerase found in T-PLL cells are sufficient to stably maintain the critically short telomeres and allow clonal expansion. In addition, we can demonstrate that inhibition of telomerase in vitro in the situation of T-PLL cells with already very short telomeres and high telomerase levels leads to rapid cytotoxicity of the T-cell clone without a time delay before telomeres get critically short as observed in the situation of longer telomeres. Targeting telomerase and telomeres seems therefore an attractive strategy for the future treatment of this devastating disease.

This study was designed to evaluate the utility of flow-FISH telomere length measurement in white blood cells (WBC) as a screening test for Dyskeratosis congenita (DC). We studied 26 patients: 17 with DC, 1 silent carrier (clinically normal; mutation in TERC), 4 with the Hoyeraal-Hreidarsson variant (HH), and 4 with Revesz Syndrome. Five had mutations in DKC1, 5 in TERC, and 2 in TERT. 23 had hematologic abnormalities, 19 had 2 or 3 of the DC diagnostic triad (lacey pigmentation, dyskeratotic nails, and leukoplakia), and 4 had soft signs of DC. We evaluated 54 first-degree relatives of DC patients, 16 Fanconi Anemia patients (FA), 14 with Diamond-Blackfan Anemia (DBA), 5 with Shwachman Diamond Syndrome (SDS), and 10 with other possibly inherited cytopenias (Other). Telomere length was measured in granulocytes and lymphocyte subsets by automated multicolor flow-FISH; results were compared with age-matched values from 400 normal controls. “Very low (VL)” telomere length was defined as a mean telomere length below the normal first percentile for age and specific WBC type. We observed VL telomeres in all subsets in the silent carrier, all HH and Revesz patients, and 15/17 with DC. Eight of 51 DC relatives had VL telomeres in granulocytes versus 2/54 with VL telomeres in lymphocytes. The sensitivities for distinguishing a DC patient from an unaffected relative were 92% in lymphocytes and 96% in granulocytes; the specificities were 96% and 98%, respectively; the sensitivity and specificity for VL telomeres in both cell types were 96% and 96%. The silent carrier with a TERC mutation developed thrombocytopenia, hypocellular marrow, and a cytogenetic clone during follow-up. The 2 DC relatives with VL telomeres in lymphocytes were from a family without a known mutant gene; they may also be silent carriers. The latter possibility disqualified an HLA-matched sibling as a donor for DC-related aplastic anemia, because of engraftment concerns; another sibling donor with normal telomere length was selected. VL granulocyte telomeres were observed in 5/16 FA, 3/14 DBA, 1/5 SDS, and 1/10 Other patients, versus 2/16 FA, 1/14 DBA, 1/5 SDS, and 0/10 Other in lymphocytes, and in both lineages in only 1 each of FA, DBA, and SDS. The sensitivity and specificity for distinguishing DC from non-DC patients using VL telomeres in both lineages were 96% and 93%, respectively. Only DC patients had consistently VL telomeres in all cell subsets. Flow-FISH telomere length measurement provides a sensitive and specific method for identifying patients with DC among families, regardless of mutation status, and distinguishes patients with DC from those with other inherited or acquired marrow failure syndromes. It may also help to detect silent carriers, and facilitate identification of mutations in other telomere biology genes. Our data suggest that the diagnostic triad, soft physical findings and/or bone marrow failure may not be required for the diagnosis of DC. Correct diagnosis of DC will enhance genetic counseling and hematologic management.

Mutations in telomere repair complex genes TERT(encoding telomerase reverse transcriptase) and TERC(telomerase RNA component) are associated with bone marrow failure, especially acquired aplastic anemia and dyskeratosis congenita. Low telomerase activity leads to short telomeres of leukocytes, predisposing highly proliferative tissues such as the bone marrow to early senescence and exhaustion of the stem cell compartment. Telomere repair gene mutations have been suggested to result in disease anticipation, defined as earlier and/or worsening clinical manifestations in successive generations. We have identified a six-generation pedigree in a large Mennonite family carrying a novel TERTmutation (K570N), which localizes in the catalytic domain with reverse transcriptase activity (RT domain). The index patient is a 26 year old male dairy farmer with a ten-year history of severe aplastic anemia (5% bone marrow cellularity) unresponsive to immunosuppression. A long history of hematologic diseases was well known and named in the family--the patient's paternal great-great-grandmother had died of a severe blood disorder at age 65 years. However, the great-grandmother and the grandfather had never presented any hematological disease. The patient's father had myelodysplastic syndrome at age 33 years, evolving to acute myeloid leukemia and death due to failure to recover blood counts after chemotherapy. One of the proband's paternal aunts had aplastic anemia develop when she was a young woman and has been transfusion-independent for decades in response to chronic androgen therapy. A second proband's paternal aunt underwent a liver transplant at age 20 for submassive hepatic necrosis with fibrosis. A third proband's paternal aunt has macrocytosis only at age 47. Two sisters (ages 21 and 23) also have macrocytosis in the absence of other hematological abnormality, and two other sisters are healthy. Genetic analysis showed that TERT K570N mutation is present in the patient's paternal (including grandfather, three aunts and two affected sisters) but not maternal relatives (making his father an obligatory carrier). The patient's oldest of three sons, now age four years, has the mutation but is asymptomatic. There was no nail dystrophy, leukoplakia or skin hyperpigmentation in any of the TERT K570N carriers; although the index patient and some of his relatives showed early graying of hair, this characteristic did not track with the mutation. Telomere shortening of leukocytes, as measured by Flow-FISH, tracked to the mutation in three generations analyzed, being shortest in the proband and in his aunt with marrow failure. Mutagenized TERTvectors transfected into telomerase-deficient VA13 cell lines yielded no telomerase activity using the telomeric repeat amplification protocol (TRAP) assay, whereas when wild-type TERTvectors were co-transfected, telomerase activity was approximately half of wild-type transfected only, indicating haploinsufficiency as a mechanism of telomere shortening. Our results confirm the association between aplastic anemia and TERTmutations. The pattern of hematologic disease in this kindred does not support disease anticipation in TERTmutations. Most remarkably, there is a likely relationship between a telomerase gene mutation and hematological malignancy and severe liver disease.

Because hematopoietic stem cells (HSC) cannot be directly observed, stochastic simulations in conjunction with competitive repopulation experiments have been used to estimate the replication rate of feline (1 rep per 8–10 wks) and murine (1 rep per 2.5 wks) HSC (model description and estimates in Nat Med2:190, 1996; Blood96:3399, 2001). These results could suggest that the HSC replication rate decreases with animal size and longevity, perhaps implying that the number of lifetime replications per HSC is limited and conserved in mammals. To test this, we analyzed retroviral vector gene marking and granulocyte telomere length data from baboons and rhesus macaques. Rhesus macaques are approximately the same size (weight 8 kg) as cats and have a similar lifespan (15–20 years); their hematopoietic demand, defined as the number of blood cells required per lifetime, is comparable. In contrast, baboons are larger (15 kg), live for 30 years, and have a hematopoietic demand more similar to humans. We simulated HSC dynamics for virtual baboons and macaques using specific HSC replication rates and then determined if observed data could be a random draw from 1000 simulated datasets. Compatibility of simulated and observed gene marking studies was determined by 3 formal criteria computed for both observed and simulated data: 1) the time after transplantation until the percent of marked cells stabilizes (measured using a change point model), 2) the presence or absence of drift after stabilization (drift is defined as a post-stabilization slope significantly different from 0 and estimated as at least 2% per 100 days), and 3) the amount of residual variation after stabilization (variation about a smoothed fit to the data -- transforming using the variance-stabilization transformation, arcsine of square root). Binomial probabilities were computed to evaluate Criterion 2 and Kolmogorov-Smirnov goodness of fit tests were used for evaluating Criteria 1 and 3, with p-values < 0.05 defining incompatibility. Mice and cat HSC parameters did not yield simulated data compatible with the observed baboon data. Rather, the analyses required that baboon HSC replicated less frequently, approximately once per 27–77 weeks (with 300 transplanted HSC (the estimated number infused); the range was robust to other modeling assumptions). In contrast, cat HSC parameters yielded simulated data compatible with the observed macaque data if 100–500 HSC were transplanted, whereas mice HSC parameters trended towards incompatibility (incompatible for 100 and 500 transplanted HSC; nearly incompatible for 300 transplanted HSC, p=0.07). Next, granulocyte telomere length data were simulated (methods in Exp Hematol32:1040; 2004) and preliminary data yielded best estimates for the replication rate of HSC in macaques and baboons of once per 18 weeks and once per 24 weeks, respectively. Our results (derived from 2 independent experimental approaches) demonstrate that the replication rate of macaque HSC is slower than mouse, whereas the replication rate of baboon HSC is slower than both cat and mouse; argue that the rate of HSC replication inversely correlates with longevity; and support Hayflick’s hypothesis that cells can only undergo a relatively constant finite number of lifetime replications. This apparent evolutionary constraint may extend to human HSC behavior.

The mechanisms underlying the aging of the hematopoietic stem cell (HSC) compartment remain poorly understood. The ATP-binding cassette cell surface transport protein, ABCG2, has been identified as the transporter responsible for Hoechst 33342 (Hst) efflux in primitive stem cells and its expression is associated with the side population (SP) phenotype in both murine and human bone marrow (BM). ABCG2 expression and Hst efflux activity is highest in those cells with the greatest repopulating potential and is progressively downregulated during differentiation. The substrate profile of ABCG2, which includes a number of antineoplastic drugs, protoporphyrin IX and the chlorophyll breakdown product pheophorbide, suggest that ABCG2 transporters may function to protect stem cells from cytotoxic insults, a function which may be of great importance in stem cell maintenance. Amongst laboratory mice, the C57BL/6 strain is the longest lived and appears to accumulate HSC’s with age as assessed by both phenotype and colony forming assays. While the phenotypic features of the SP profile have been well characterized in both humans and young mice, little is known of the Hst efflux dynamics or phenotype of the SP profile in old and very old C57BL/6 mice. In order to further characterize the SP profile in old mice, whole BM was extracted from the femurs, tibiae, pelvis and thoracolumbar vertebral bodies of young (9–13 week) and old (95–108 week) C57BL/6 (Ly5.1) mice. Cells were stained with 5μg/ml Hst followed by staining with a combination of CD45.1 FITC, Sca1 PE, c-kit APC, CD34 FITC, biotinylated CD34 and lineage markers and strep PE-Texas Red. In addition, serial sampling of Hst incubated cells was performed to assess Hst efflux activity at 20 mins incubation through to 100mins. Six-color flow analysis was performed on a FACS Vantage™ (BD) cytometer and data analyzed using FlowJo™ software. Results show a marked increase in cells with an SP phenotype in old vs young mice (mean±SD 1.85%±0.88 vs 0.15%±0.09) which were more highly enriched for CD34-Sca1+ckit+ (22.2%±8.65 vs 8.89%±6.7) cells. Subdividing the SP profile into four regions (R1 to R4) from highest to lowest Hst efflux activity revealed that in old mice, SP cells with the highest Hst efflux activity were almost exclusively of a CD34-Sca1+ckit+ phenotype (82.3%±14.0 vs 11.5%±7.8), with a decreasing proportion of these cells represented throughout the remaining SP tail, though a significant proportion of cells within R4 remain CD34-Sca1+ckit+ (15.3%±7.83 vs 4.19±3.01). Similar patterns have been observed in both whole and lineage depleted BM. In addition, BM cells from old C57BL/6 mice show prolonged Hst efflux activity with an increase in cells in the SP gate at 100 mins (1.51%±0.50 vs 0.10%±0.06). We conclude that in old C57Bl/6 mice, cells accumulate which have the capacity to efflux Hst in agreement with previous reports of an increase in HSC number with age in this mouse strain.

Telomere shortening is observed in some patients with acquired aplastic anemia (AA) and associates with poor response to immunosuppressive therapy. Mutations in components of the telomerase complex (RNA component, TERC, and reverse transcriptase, TERT) have been found in AA patients. As the proportion of AA patients with short telomeres is greater than the number of patients with telomerase complex component mutations, we investigated whether genetic variations in the telomere-binding proteins telomeric repeat binding factor 1 (TERF1), telomeric repeat binding factor 2 (TERF2) and TERF1-interacting nuclear factor 2 (TIN2) were also associated with AA. TERF1 is a negative regulator of telomere length; TIN2 regulates TERF1 function, and TERF2 protects telomeres from degradation and fusion. Bi-directional sequence analysis was performed across all exons and proximal promoter regions of TERF1, TERF2, and TINF2 genes in 147 patients with AA and 118 healthy subjects. Haplotypes were inferred by Phase 2.0. Telomere length of leukocytes was measured by flow cytometry fluorescent in situ hybridization (flow-FISH). A nonsynonymous mutation in TERF1, exon 9, codon 377 (Ala → Val) was found in a 15-year old African-American girl who did not respond to immunosuppressive therapy, but not in controls. A specific TERF1 haplotype was more common in patients (P=0.004), whereas another haplotype was more common in controls (P=0.02). For TERF2 , a mutation in exon 6, codon 273 (Ala→Ser) was found in a 40 year-old Caucasian female patient, but not in controls. A G/A single nucleotide polymorphism (SNP) in exon 6 was present in 15% of patients and 9% of controls (P=0.04). TINF2 was the most polymorphic gene in both patients and contorls. As these variants were found mainly in African-Americans, additional 94 healthy African-American controls were sequenced. In the proximal promoter region, two SNPs and one 16-base-pair deletion were found in patients, but not in 212 controls. The 16-base-pair deletion (−10 to −25) was found in a 31-year-old African-American female who did not respond to immunosuppression, with very short telomeres of leukocytes (Δ TRF −2.59 kb). Also, a -97G/C transition was found in a 66-year-old African-American female who partially responded to immunosupression, also tracking to short telomeres (ΔTRF, −1.22 kb). The −260G/C promoter SNP was present in 3.5% of controls and 0.4% of patients P<0.0001), and the −91C/T SNP was seen in 1.7% of controls but not in patients (P<0.0001). The most common haplotype was seen in 94% of patients and 86% of controls (P=0.003), suggesting an at risk haplotype. A rare haplotype in patients (0.4%) was more common in controls, 3.4% (P=0.01) suggesting a rare but potentially protective variant. These data suggest that mutations and specific haplotypes may confer risk for AA. In addition, we have identified novel mutations in three genes that are essential for telomere stability. In conclusion, mutations in different genes involved in telomere stability and repair appear to be hematologic genetic risk factors for bone marrow failure in adults.

About one third of patients with acquired aplastic anemia (AA) have very short telomeres of their peripheral blood leukocytes; short telomeres correlate to long disease duration and poor response to immunosuppressive therapy. We have recently shown that some cases of apparently “acquired” AA have mutations in TERC, the gene encoding the RNA component of telomerase (Fogarty et al., Lancet2003;632:1628; Yamaguchi et al., Blood2003;102:916). In the current study, we examined TERT, the gene encoding the telomerase reverse transcriptase, by sequencing the gene’s exons and proximal promoter region in peripheral blood DNA samples from 122 patients with AA and 282 controls. Four novel nonsynonymous mutations among five patients, not present in controls, were discovered; three polymorphisms were identified, two nonsynonymous SNPs and one deletion of a single amino acid. To investigate the functional consequences of the mutations, telomere lengths of leukocytes were assessed by flow cytometric fluorescence in situ hybridization (flow-FISH). All patients carrying TERT mutations had markedly short telomeres compared to age-matched controls, as opposed to other AA patients with polymorphisms, whose telomeres were normal. In one patient’s kindred, presence of the TERT mutation in other family members correlated to telomere shortening; non-carriers had normal telomeres. Telomerase function in patients’ T cells, activated by phytohemagglutinin and interleukin-2 to increase enzymatic activity, was measured by the telomeric repeat amplification protocol (TRAP). In all mutant AA patients evaluated, cell lysates yielded very low or no telomerase activity, as compared to normal controls. We cloned the AA-related mutations into a TERT expression vector and co-transfected these vectors into VA13 cells (with a TERC-containing vector, as this cell line does not have telomerase activity due to absent TERC and TERT expression). All mutant TERT-containing cell lysates were severely deficient in enzymatic activity. TERT gene expression, as evaluated by Northern blot, was normal in cells transfected with the mutated genes. When vectors containing wild-type TERT and individual TERT mutations were cotransfected, telomerase activity was dramatically reduced, in comparison to transfection of wild-type TERT vector only. These results indicate a dominant negative mechanism of action of TERT mutations as responsible for the absence of telomerase activity in AA patients’ cells. Family members lacking telomerase activity have short telomeres but appear physically normal and have no hematological abnormalities. In a provisional model of the telomere repair complex and marrow failure, mutations in DKC1 and the stability regions of TERC cause dyskeratosis congenita, with early presentation and associated physical anomalies. Mutations that affect the enzyme-binding region of TERC and in TERT, the reverse transcriptase itself, lead to a constitutionally reduced stem cell compartment and appear to be genetic risk factors for the development of “acquired” aplastic anemia.

Human telomerase uses a portion of its integral RNA component (hTER) as the template to synthesize telomeres at chromosome ends. hTER sequence polymorphisms have been observed in some patients with bone marrow failure syndromes such as aplastic anemia, but the functional significance of most such variants is unknown. Here, we report the functional characteristics of ten previously-described and two newly discovered hTER disease-associated polymorphisms. Most of these hTER variants adversely affected telomerase enzymatic function as measured in the telomerase reconstituted human cells. Similar loss-of-function effects were also seen directly in primary lymphocytes collected from two of the patients. The majority of the functional deficits stemmed from perturbations of the predicted hTER RNA secondary structure, and corresponded well with the degrees of telomere shortening observed in patients. In contrast, hTER variants anticipated to be inconsequential polymorphisms, which were also found in healthy subjects, did not interfere with telomerase function. Loss of telomerase activity and of telomere maintenance resulting from inherited hTER mutations may predispose some patients to aplastic anemia and other marrow failure disorders.

This unit describes a quantitative technique for measuring the lengths of telomere repeat sequences in individual chromosomes from single metaphase cells. The technique is based on fluorescence in situ hybridization (FISH) adapted for use with peptide nucleic acid (PNA) probes. PNA is an example of novel synthetic oligonucleotide “mimetic” which has a higher affinity than regular oligonucleotide (RNA or DNA) probes for complementary single‐strand (ss) DNA sequences. PNA oligonucleotides have excellent penetration properties due to their small size (typically 15 to 18‐mers) and can be directly labeled with fluorochromes. These properties have been exploited to develop quantitative fluorescence in situ hybridization (Q‐FISH) onto denatured single‐stranded chromosomal DNA target sequences. The latter can be present in preparations of fixed metaphase cells on slides (Q‐FISH) or in heat‐treated (interphase) cells in suspension (flow‐FISH).