Heterogeneous telomere defects in patients with severe forms of dyskeratosis congenita

Heterogeneous telomere defects in patients with severe forms of dyskeratosis congenita Fabien Touzot, MD, PhD,a,b,c,i Laetitia Gaillard,a,b,i Nadia Vasquez,d Tangui Le Guen, MSc,a,b,i Yves Bertrand, MD, PhD,e,f Jean Bourhis, MD, PhD,g Thierry Leblanc, MD, PhD,h Alain Fischer, MD, PhD,a,b,c,i Jean Soulier, MD, PhD,d Jean-Pierre de Villartay, PhD,a,b,c,i and Patrick Revy, PhDa,b,i Paris, Lyon, and Villejuif, France Background: Telomeres represent the tips of linear chromosomes. In human subjects telomere maintenance deficiency leads to dyskeratosis congenita (DC), a rare genetic disorder characterized by progressive bone marrow failure, accelerated aging, and cancer predisposition. HoyeraalHreidarsson syndrome (HH) is a severe variant of DC in which an early onset of bone marrow failure leading to combined immunodeficiency is associated with microcephaly, cerebellar hypoplasia, and growth retardation. Objectives: Limited information is available on the cellular and molecular phenotypes of cells from patients with HH. We analyzed fibroblasts and whole blood cells from 5 patients with HH, 3 of them of unknown molecular origin. Methods: Telomere length, cellular senescence rate, telomerase activity, telomeric aberration, and DNA repair pathways were investigated. Results: Although patients’ cells exhibit dysfunctional telomeres, sharp differences in the telomeric aberrations and telomere lengths were noted among these patients. In some patients the dysfunctional telomere phenotype was unprecedented and associated with either normal telomere length or with telomeric aberrations akin to fragile telomeres. This result is of particular importance because the molecular diagnosis of these patients is primarily based on telomere length, which therefore misses a subset of patients with telomere dysfunction. Conclusion: These observations provide the notions that (1) various telomere defects can lead to similar clinical features, From aINSERM, U768, Paris; bFacult
e de M
edecine Ren
e Descartes, Universit
e Paris Descartes, Site Necker, IFR94, Paris; cAP-HP, H^opital Necker Enfants-Malades, Service d’Immunologie et d’H
ematologie P
ediatrique, Paris; dINSERM U944, Institut universitaire d’H
ematologie, Universit
e Denis Diderot, Paris; eInstitut d’h
ematologie et d’oncologie p
ediatrique, Lyon; fUniversit
e Claude Bernard Lyon I, Lyon; gUPRES EA 27-10 Institut Gustave Roussy, Villejuif; hAP-HP, H^opital Robert Debr
e, Service d’h
ematologie p
ediatrique, Paris; and iFondation Imagine Paris. Supported by institutional grants from the Institut National de la Sant
e et de la Recherche M
edicale, Ligue Nationale contre le Cancer (Equipe Labellis
ee La Ligue), Association pour la Recherche sur le Cancer, Institut National du Cancer (INCa)/Canc
erop^ole Ile de France, DHOS (centre de R
ef
erence Maladies Rares ‘‘Aplasies m
edullaires constitutionnelles’’), and the European Research Council. F.T. received fellowships from the Fondation pour la Recherche M
edicale, and T.L.G. received fellowships from La Ligue. P.R. is a scientist from the Centre National de la Recherche Scientifique (CNRS). Disclosure of potential conflict of interest: A. Fischer has received research support from the European Research Council and INSERM. The rest of the authors declare that they have no relevant conflicts of interest. Received for publication July 7, 2011; revised September 14, 2011; accepted for publication September 22, 2011. Available online November 10, 2011. Corresponding author: Patrick Revy, PhD, INSERM U768, 149 rue de Sevres, 75015 Paris, France. E-mail: [email protected]. 0091-6749/$36.00 Ó 2011 American Academy of Allergy, Asthma & Immunology doi:10.1016/j.jaci.2011.09.043

(2) telomere dysfunction in cells from patients with DC/HH is not always associated with short telomeres, and (3) additional factors, likely involved in telomere protection rather than in length regulation, are responsible for a subset of DC/HH. (J Allergy Clin Immunol 2012;129:473-82.) Key words: Telomere, dyskeratosis congenita, Hoyeraal-Hreidarsson syndrome, senescence, immune deficiencies, bone marrow failure

All dividing cells with linear chromosomes face challenges inherent to the particular structure of chromosome ends: the telomeres. One of these challenges is the inability for the DNA replication machinery to fully replicate the tip of the chromosome.1 This would result in the progressive loss of telomeric sequence on cell divisions. Telomerase, a ribonucleoprotein complex including telomerase reverse transcriptase (TERT), telomerase RNA component (TERC), Dyskerin, NOP10, and NHP2, counteracts the cell cycle–dependent telomere attrition by adding telomeric sequences. However, the expression of TERT, the catalytic subunit of telomerase, is restricted to germ cells, stem cells, and some activated cells.2 Consequently, after many divisions, somatic cells harbor short telomeres leading to cell-cycle arrest, a phenomenon known as replicative senescence.3 Telomeres exhibit a unique structure consisting of a long G-rich strand overhang that invades the duplex telomeric repeat to form a telomeric loop (T-loop). This structure, maintained by a specific telomeric complex called the shelterin, impedes the telomere from being recognized and processed as DNA-double strand breaks. Telomeric repeat binding factor 1 (TRF1), TRF2, RAP1, protection of telomeres protein 1 (POT1), TPP1, and TRF1-interacting protein 2 (TIN2) are the 6 known subunits of the shelterin complex.4 They are dedicated to protecting chromosome ends from degradation, fusion, or both, as well as regulating the recruitment and activity of the telomerase complex.1 In addition, the replication of DNA at telomeres is a process that is particularly challenging for proliferative cells. Indeed, the unique structure and DNA repeats that form telomeres require cooperation between the replication machinery, telomeric factors, and several DNA damage-response components, including the DNA repair factor Apollo, in a process that is not fully understood. All of these molecules, acting in a concerted manner, are required to properly replicate telomeres and thus avoid genomic instability.5 In human subjects defective telomere maintenance is responsible for the condition dyskeratosis congenita (DC). DC is a rare disorder characterized by progressive bone marrow failure, accelerated aging, mucocutaneous abnormalities, and cancer predisposition.6-8 The Hoyeraal-Hreidarsson syndrome (HH) is a severe variant of DC in which early-onset bone marrow failure is associated with microcephaly, cerebellar hypoplasia, and 473

474 TOUZOT ET AL

J ALLERGY CLIN IMMUNOL FEBRUARY 2012

Antibodies Abbreviations used CO-FISH: Chromosome orientation–fluorescence in situ hybridization DC: Dyskeratosis congenita DSB: Double-strand break HH: Hoyeraal-Hreidarsson syndrome ICL: Interstrand crosslink IRIF: Ionizing radiation–induced foci ITS: Interstitial telomeric signal MMC: Mitomycin C POT1: Protection of telomeres 1 TERC: Telomerase RNA component TERT: Telomerase reverse transcriptase TIF: Telomere dysfunction–induced foci TIN2: TRF1-interacting protein 2 TRAP: Telomeric repeat amplification protocol TRF: Telomeric restriction fragment

Rabbit polyclonal anti-53BP1 was from Novus Biologicals (Littleton, Colo), and Alexa488 goat F(Ab)92 secondary antibodies were from Molecular Probes (Eugene, Ore).

Immunofluorescence detection Immunofluorescence experiments were performed as described by Touzot et al.17

Senescence-associated b-galactosidase staining Primary fibroblasts were fixed for 5 minutes in 4% vol/vol paraformaldehyde in PBS, washed in PBS, and stained in b-galactosidase fixative solution (X-gal) in 5 mmol/L potassium ferricyanide, 5 mmol/L potassium ferrocyanide, and 2 mmol/L MgCl2 in PBS for 18 hours at 378C.

Ionizing radiation–induced foci and mitomycin C sensitivity assays intrauterine growth retardation.7,9 Patients with HH generally succumb before 10 years of age, underlining the severity of this syndrome. Mutations in the telomerase factors TERT, TERC, Dyskerin, TCAB1, NOP10, NHP2, and the shelterin TIN2 were reported in patients with DC/HH, linking impaired telomere maintenance to the disease.10,11 However, the molecular cause of about 50% of the cases of DC/HH remains undetermined. Because whole blood cells from most of the patients with DC/HH studied thus far exhibit abnormally short telomeres, DC and HH are thought to be caused by accelerated telomere shortening.7 In addition, mice in which telomere length is artificially reduced exhibit symptoms of accelerated aging, further supporting the association between a short telomere and a degenerative defect, as observed in patients with DC/HH.12-14 Conversely, increasing telomere length can result in healthy aging and longevity in human subjects and mice.15,16 It is therefore accepted that the symptoms observed in patients with DC/HH are caused by excessive telomere shortening, mainly affecting the highly dividing stem cells. To further delineate the molecular events leading to severe form of DC, we analyzed fibroblasts from 5 patients with HH (1 TIN2-deficient patient, 1 Apollo-deficient patient,17 and 3 patients with yet uncharacterized molecular origin). The fibroblasts from these patients with HH exhibited distinct telomere dysfunctions that were not systematically associated with short telomeres.

METHODS Patients and cells In accordance with the Helsinki Declaration, informed consent for our study was obtained from the families. This study was also approved by the INSERM Institutional Review Board. The whole blood cell DNA used as short telomere control (TIN2 1/2) was obtained from a boy harboring a de novo heterozygous mutation in the TINF2 gene (c.849delC p.T284GfsX33). SV40-transformed and telomerase-immortalized cell lines were obtained as previously described.18 HH, as defined by Hoyeraal et al19 and Hreidarsson et al20 and reviewed by Ohga et al,21 is characterized by intrauterine growth retardation, microcephaly, mental retardation, cerebellar malformation, progressive bone marrow failure, and mucocutaneous lesions. The 5 patients with HH here described meet the defining criteria of this syndrome.

Primary fibroblasts were seeded on cover slips and x-ray irradiated. One or 24 hours after irradiation, cells were fixed and incubated with appropriate antibodies, as described by Buck et al.18 The flow-based mitomycin C (MMC) sensitivity assay was performed, as described by Pinto et al.22

Chromosome orientation–fluorescence in situ hybridization and image capture Chromosome orientation–fluorescence in situ hybridization (CO-FISH) and image capture were performed as previously described.23

Telomeric restriction fragment Measurement of the lengths of the telomeric restriction fragments (TRFs) was performed by using Southern blotting, as described by Touzot et al.17

Statistical analysis For telomeric aberrations observed by using CO-FISH, statistical analysis was conducted as previously described.17 Telomeric aberrations of each condition were compared considering the total number of chromatid ends. For analysis of ionizing radiation–induced foci (IRIFs) disappearance 24 hours after irradiation, P values were obtained by using the Student t test.

RESULTS Clinical features of patients Five patients (HH1 [previously reported by Touzot et al17], HH2, HH3, HH4, and HH5) presented with features compatible with the severe form of DC known as HH (see the Methods section for the definition of HH). All 5 patients presented with microcephaly, facial dysmorphia, or both; cerebellar hypoplasia; and intrauterine growth retardation and had bone marrow failure at the mean age of 15.6 months (range, 8 months to 2.5 years; Table I).9 Of note, some of the patients exhibited profound B-cell lymphopenia. HH1 had agammaglobulinemia with B alymphocytosis,9 HH2 had moderate hypogammaglobulinemia (IgG, 4.63 g/L [normal, >4.8 g/L]; IgA, 0.25 g/L [normal, >0.3 g/L]; and IgM, 0.29 g/L [normal, >0.5 g/L]) with B-cell lymphopenia (160 lymphocytes/mm3 [normal, >360 lymphocytes/mm3). HH3 had the classical mucocutaneous features of DC (Table I). Three patients (HH3-HH5) received hematopoietic stem cell transplantation with matched unrelated donor or matched cord blood after a myeloablative conditioning regimen. Patients HH3

TOUZOT ET AL 475

J ALLERGY CLIN IMMUNOL VOLUME 129, NUMBER 2

TABLE I. Clinical features of patients with HH Sex

Cons

Age at diagnosis

Hematologic features

Developmental features

HH1

F

No

18 mo

Aplastic anemia IUGR B alymphocytosis Cerebellar Agammaglobulinemia hypoplasia Microcephaly

Severe enteropathy Thin and sparse hair

Death at 4 y old after severe infection

Expression of a unique Apollo variant exerting a dominant negative effect on telomere protection17

HH2

F

Yes

2.5 y

Aplastic anemia B lymphopenia Hypogammaglobulinemia

IUGR Cerebellar hypoplasia Microcephaly

Death at 3.5 y old after pulmonary infection

Unknown

HH3

F

No

8 mo

Aplastic anemia

IUGR Cerebellar hypoplasia

Enophtalmia Bulging forehead Broad fingers Ligamentous hyperlaxity Nail dystrophy Leukoplakia

HH4

M

No

12 mo

Aplastic anemia

HH5

F

No

10 mo

Aplastic anemia with Myelodysplastic features

IUGR Cerebellar hypoplasia Microcephaly IUGR Cerebellar hypoplasia Microcephaly

Other features

Nail dystrophy Leukoplakia

Nail dystrophy Peripheral demyelinating neuropathy

Outcome

Molecular origin of the disease

HSCT with cord blood Unknown Pulmonary fibrosis Chronic GVHD Death at 2 y old after GVHD Death after HSCT with MUD Heterozygous de novo Chronic GVHD TIN2 mutation (K280X) Pulmonary and hepatic fibrosis B-ALL with monosomy 7 Unknown Toxic death after HSCT with MUD at 2 y old

B-ALL, B-lineage acute lymphoblastic leukemia; Cons, consanguinity; F, female; GVHD, graft-versus-host disease; HSCT, hematopoietic stem cell transplantation; IUGR, intrauterine growth retardation; M, male; MUD, matched unrelated donor.

and HH4 had severe graft-versus-host disease and organ failure, including pulmonary fibrosis and cirrhosis likely related to HH (Table I). All patients died of severe infection or after the transplantation procedure. Of note, patient HH2 was born to consanguineous family, suggesting a possible autosomal recessive inheritance of the disease. Interestingly, patient HH5’s father was healthy, whereas the mother had progressive ataxia and dysarthria that were evaluated at the age of 28 years by means of magnetic resonance imaging, showing a profound cerebellar atrophy. She had, like her daughter, a demyelinating sensorimotor peripheral neuropathy. This feature evokes disease anticipation, in which the onset of symptoms is earlier and more severe in the children than in the parents, as previously described in families with autosomal dominant DC caused by a defect in TERC and TERT.24-26 This suggests that HH in patient HH5 might be of autosomal dominant inheritance.

Genetic analysis of patients with HH Sequencing analysis of the known HH-causing genes revealed a de novo heterozygous TINF2 mutation, c.838 A>T (cDNA), in patient HH4 leading to the truncation of the TIN2 protein at amino acid position 280 (p.280 K>X; see Fig E1 in this article’s Online Repository at www.jacionline.org). The same TINF2 mutation has been previously reported in 1 patient with HH,27 but the consequences of this mutation on cellular phenotype were not previously addressed. In addition, in patient HH1 we identified a unique splice variant of the DNA repair factor Apollo. The dominant negative effect exerted by this Apollo variant on telomere protection in patient HH1 has recently been reported.17 Apollo cDNA from primary fibroblasts of the other patients was amplified and sequenced. No splice variant or mutation in Apollo transcript were detected. Exon sequencing of the other known HH-causing genes (TERC, TERT, TINF2, Dyskerin, NOP10,

NHP2, and TCAB1), as well as the shelterin genes (TRF1, TRF2, POT1, TPP1, and RAP1) did not reveal any mutation in patients HH2, HH3, and HH5.

Primary fibroblasts from patients with HH exhibit dysfunctional telomeres Primary fibroblasts from patient HH1, as previously described,17 were used as controls of telomere dysfunction. Fibroblast cell lines from patients HH2 to HH5 were generated, and cellular senescence was evaluated at early passage through the detection of senescence-associated b-galactosidase activity (see Fig E2, A, in this article’s Online Repository at www.jacionline.org). The proportion of senescent cells, although variable among patients, was sharply higher (range, 31% to 74%; P < .001) than that of primary fibroblasts from healthy donors with higher passage numbers (<2%; passage 11; Fig 1, A). In addition, the detection of a DNA damage response specifically localized at telomeres (known as telomere dysfunction–induced foci [TIFs]; see Fig E2, B)28 was significantly higher in all cells from patients with HH (P <.0001). Although 70% of control primary fibroblasts did not exhibit TIFs, this proportion was consistently less than 30% in patients’ cells (Fig 1, B). In addition, 10% to 27% of primary fibroblasts from patients with HH exhibited 4 or more TIFs per cell, whereas control fibroblasts were devoid of such signal (Fig 1, C). Collectively, the combined detection of robust DNA damage response at telomeres and increased cellular senescence indicate that primary fibroblasts from these patients with HH manifest hallmarks of telomere dysfunction. Functional telomerase complex in cells from patients with HH Most of the cases of DC/HH described to date are caused by defective components of the telomerase complex (ie, TERT,

476 TOUZOT ET AL

J ALLERGY CLIN IMMUNOL FEBRUARY 2012

FIG 1. Telomere dysfunction in primary fibroblasts from patients with HH. A, Cellular senescence. Results are expressed as the percentage of senescence-associated b-galactosidase (SA-b-Gal)–positive primary fibroblasts. Cells from patients with HH exhibit a high rate of senescence compared with primary fibroblasts from a healthy donor. At least 100 cells were analyzed by condition. B, Quantification of TIFs. The number of TIFs per nucleus (materialized by the colocalization of 53BP1 with TRF2, see Fig E2) was quantified. The number of cells analyzed is indicated in Fig 1, C. C, The percentages of fibroblasts from a control subject (n 5 204) and the patients with HH (HH1, n 5 116; HH2, n 5 120; HH3, n 5 139; HH4, n 5 85; and HH5, n 5 89) presenting with 4 or more TIFs are indicated. The cell passage numbers are indicated in brackets (P). Ctl, Control subject.

TERC, Dyskerin, NOP10, and NHP2) that result in impaired telomerase activity,6,29 In vitro telomerase activity was assessed by using the telomeric repeat amplification protocol (TRAP)30 with extracts from SV40-transformed fibroblasts transduced with a human TERT–expressing vector (fibroblasts do not express human TERT). Because of a lack of material, TRAP was not evaluated in patient HH3 and the TIN2-deficient (patient HH4) cells. One microgram of protein extracts from wild-type cells and cells from patients HH1, HH2, and HH5 was sufficient to reveal telomerase activity (Fig 2, A, lanes 1, 7, 10, and 13, respectively). Conversely, TRAP activity in Dyskerin-deficient cells, which were used as a telomerase defective control, was only observed with 10 mg of extracts (Fig 2, A, lanes 4-6), as expected.29 This result indicates that the telomere dysfunction observed in fibroblasts from patients HH1, HH2, and HH5 is caused neither by impaired TERC expression nor by a defect of a factor involved in TERC stability.

Fibroblasts from patients with HH do not inevitably exhibit short telomeres Cells with too short telomeres display TIFs and senescence,31 and DC/HH conditions are commonly considered excessive telomere-loss diseases.10 Given the absence of impaired telomerase activity noted above, we next evaluated the telomere length in fibroblasts from patients with HH. DNA from primary fibroblasts at early passage was extracted and submitted to telomeric TRF analysis, combining DNA digestion and Southern blotting with a specific telomeric probe. Consistent with reports describing a sharp telomere length reduction in whole blood cells from TIN2-deficient patients,27,32 the TIN2-deficient primary fibroblasts (patient HH4) exhibited excessively short telomeres (mean telomere length, 3.9 kb; passage 4; Fig 2, B). Similarly, primary fibroblasts from patient HH3 exhibited abnormally short telomeres (mean telomere length, 4.6 kb; passage 4). Strikingly, the mean telomere lengths in primary fibroblasts from patients HH1, HH2, and HH5 (6.8, 8.6, and 8.6 kb, respectively) were similar to those of 4 control subjects (control subjects 1, 2, 3, and 4: 6.6, 6.8, 7.6, and 10.1 kb, respectively) and in the normal range.33 Knowing that human telomeres from primary fibroblasts are considered critically short when less than 5 kb,34 we concluded from our

observations that even though fibroblasts from patients with HH exhibit dysfunctional telomeres, this phenotype is not necessarily associated with excessive shortening of telomeres.

Functional DNA double-strand break and interstrand crosslink repair in fibroblasts from patients with HH Several DNA repair factors participate in telomere protection.34,35 Moreover, a defect of DNA repair factors involved in the repair of DNA interstrand crosslink (ICL) and double-strand break (DSB) causes diseases (eg, Fanconi anemia, Cernunnos, DNA-Ligase IV, or Nijmegen breakage syndrome deficiency) that share clinical features with HH. This includes bone marrow failure/lymphopenia, growth retardation, and facial dysmorphia/ microcephaly.36 The efficiency of DNA DSB and ICL repair was therefore evaluated in primary fibroblasts from patients with HH. On x-ray irradiation, DNA DSB induces the phosphorylation of the histone variant H2AX and the recruitment of DNA repair factors, such as 53BP1, at the site of the damage, forming IRIFs (see Fig E3 in this article’s Online Repository at www.jacionline.org). 53BP1 IRIFs were similarly detected in fibroblasts from patients with HH and control subjects 1 hour after 5 Gy x-ray irradiation (Fig 3, A, and see Fig E3). Twenty-four hours after DSB induction, 53BP1 IRIFs were no longer detectable in fibroblasts from control subjects and patients with HH (as opposed to Cernunnos- and ataxia telangiectasia mutateddeficient cells used as DNA-repair defective controls), indicating an efficient DNA DSB repair (Fig 3, A, and see Fig E3). Of note, the number of 53BP1 foci per nucleus in the absence of any treatment was significantly higher in cells from patients with HH (except for cells from patient HH5, P < .05) than cells from control subjects (Fig 3, A, and see Fig E3), likely reflecting spontaneous DNA damage accumulation. The ability of primary fibroblasts to repair MMC-induced ICL, a process involving the Fanconi anemia pathway, was assessed by using a flow-based assay.22 Fibroblasts from patients with HH submitted to increasing doses of MMC did not exhibit a defect in ICL repair, contrasting with fibroblasts from a patient with Fanconi anemia used as a sensitive control (Fig 3, B). Together, these

TOUZOT ET AL 477

J ALLERGY CLIN IMMUNOL VOLUME 129, NUMBER 2

FIG 2. TRAP activity and telomere length of fibroblasts from patients with HH. A, In vitro telomerase activity was analyzed by using the TRAP assay with cell extracts (1, 5, and 10 mg) from SV40–human TERT fibroblasts from a healthy donor (Ctl), a Dyskerin-deficient patient, and patients HH1, HH2, and HH5 (TRAP assay could not be performed with cells from patients HH3 and HH4 because of a lack of material). As expected, little or no telomerase activity was detected with extracts from a healthy donor treated with RNase (Ctl 1 RNase; lane 18), with extracts from primary fibroblasts (PF; lane 17), or without extract (H20; lanes 16 and 19). B, The mean telomere length of primary fibroblasts from patients HH1, HH2, HH3, HH4, and HH5 at the same passages as in Fig 1 and 4 healthy control subjects (Ctl1, Ctl2, Ctl3, and Ctl4) was estimated by using the TRF method. The estimated mean telomere length is indicated.

results indicate that telomere dysfunction in cells from patients with HH is the consequence of neither a DSB repair defect nor a ICL defective repair pathway.

Telomeric aberrations in fibroblasts from patients with HH Next we sought to know whether dysfunctional telomeres observed in fibroblasts from patients with HH could be associated with telomeric instability. Using specific fluorescent telomeric probes, we performed CO-FISH on metaphase spreads of SV40-transformed fibroblasts from patients with HH in their earliest passages. The CO-FISH technique allows us to distinguish G-rich (leading) from C-rich (lagging) telomere strands (Fig 4, A).17,37 Because of a lack of material, patient HH3 was not analyzed. Cell lines from patients with HH exhibited significant increases in telomeric aberrations, the nature of which was clearly different among patients (Fig 4, D). Cells from patient HH1 exhibited a significant increase in telomere-telomere fusions and duplications of telomere FISH signals (known as telomere doublet) without a strand preference, likely reflecting aberrant telomeric replication, as previously described.17 The telomeric FISH signal from the TIN2-deficient fibroblasts (patient HH4) was very weak, which is consistent with the short telomeres measured by TRF experiments in these cells (Fig 2, B). Accordingly, the telomeric aberrations detected in cells from patient HH4 were mostly telomere-free ends and telomere-telomere fusions (P <.001; Fig 4, D), as described for short and consequently uncapped telomeres.38 Cells from patient HH2, which had normal telomere length (Fig 2,

B), exhibited a significant increase in telomere doublets (P < .002; Fig 4, B and D). However, unlike cells from patient HH1,17 they were more frequently found on the G-rich strand (green probe) derived from leading-strand DNA replication. In addition, patient HH2 presented with frequent telomere-free ends (P <.001) without strand preference (Fig 4, B and D). The most striking telomeric aberrations and, to our knowledge, unprecedented in human cells from a patient with DC/HH were observed in fibroblasts from patient HH5. They include a significant increase (P < .001) in telomere doublets, multiple telomeric signals,39 and interstitial telomeric signals (ITSs; Fig 4, C and D). Interestingly, the ITSs and telomere doublets detected in cells from patient HH5 affected predominantly the G-rich strand (P <.05; Fig 4, D). The abundance of ITSs in cells from patient HH5, which are absent in cells from other patients with HH and control subjects, suggests a profound telomeric replication defect that mainly affects the G-rich strand. Collectively, the CO-FISH analysis provides evidence that cells from the patients with HH presented with different telomeric aberrations, some of which were unprecedented in human situations and likely resulted from severe telomeric instability. This result further supports the notion that the HH phenotype is the consequence of a profound defect in telomere physiology.

Whole blood cells from patients with HH do not invariably exhibit short telomeres Because HH is primarily a disease of the hematopoietic system, we wish to consider whether the telomeric defects seen in

478 TOUZOT ET AL

J ALLERGY CLIN IMMUNOL FEBRUARY 2012

FIG 3. DNA DSB and ICL repair in primary fibroblasts from patients with HH. A, DNA DSB repair efficiency was determined by scoring the number of 53BP1 foci in primary fibroblasts from 2 healthy donors (Ctl1 and Ctl2), a Cernunnos-deficient patient, an ataxia telangiectasia mutated-deficient patient, and the 5 patients with HH left untreated or 1 and 24 hours after 5 Gy x-ray irradiation. At least 50 cells were scored by condition. Statistical significance was determined by using the Student t test. B, Survival of fibroblasts after MMC treatment. Graphs represent the MMC sensitivity data in primary fibroblasts from a healthy control subject, a patient with Fanconi anemia (FA), and the patients with HH. On the x-axis are plotted the MMC concentration at which dying cells were detected by using propidium iodide (PI) intracellular uptake (arrows). Red and black lines represent PI uptake of untreated and MMC-treated cells, respectively.

fibroblasts resulted in abnormal telomere lengths in blood cells. Whole blood cells from 2 TIN2-deficient patients (patient HH4 and TIN2 1/2) exhibited extremely short telomere length (TRF of 4.7 and 3.5 kb, respectively; Fig 5). The same was true also in whole blood cells from patient HH3 (3.7 kb), which is in accordance with the telomere attrition seen in primary fibroblasts from patient HH3 (Fig 2, B). In contrast, the telomere lengths in whole blood cells from patients HH1, HH2, and HH5 were comparable with those observed in 15 healthy pediatric control subjects (Fig 5) and in the normal range.27 These results provide the unexpected evidence that a subset of patients with a severe form of

DC caused by telomere dysfunction can exhibit normal telomere length in primary fibroblasts, as well as in whole blood cells.

DISCUSSION In this report we analyzed the phenotype of fibroblast cell lines from 5 patients presenting with the clinical features of the severe form of DC known as HH. Because of the early deaths of these patients, no phenotypic investigation of their hematopoietic cells has been conducted. DC and HH are thought to be caused by accelerated telomere attrition because patients with DC/HH

TOUZOT ET AL 479

J ALLERGY CLIN IMMUNOL VOLUME 129, NUMBER 2

FIG 4. Telomeric aberrations in cells from patients HH2, HH4, and HH5 revealed by means of CO-FISH. A, Scheme of the CO-FISH principle. During 1 cell cycle, newly synthesized DNA strands incorporate Bromodeoxyuridine (BrdU) and Bromodeoxycytidine (BrdC). Treatment with UV irradiation and ExoIII degrades the newly synthesized strands. FISH with telomeric strand-specific probes allows us to distinguish the lagging and leading strands. B, Telomeric CO-FISH on metaphase spreads of SV40 fibroblasts from patient HH2 (passage 5) revealed an increase in telomere doublets preferentially detected by the G-rich probe. White arrows highlight the chromatid ends with telomere doublets. C, Telomeric CO-FISH on metaphase spreads of SV40 fibroblasts from patient HH5 (passage 4) revealed an increase in telomeric aberrations. Yellow arrows highlight the ITS, and red arrows highlight multiple telomeric signals (MTS), 2 features that suggest telomeric fragility during replication. D, Quantification of telomeric aberrations observed in a control subject (Ctl), HH2, HH4, and HH5. TSCE, Telomeric sister chromatid exchange. Values of telomeric aberrations in patients with HH that are statistically different from those obtained in control cells are represented in boldface.

consistently exhibited abnormally short telomeres.6,7 Moreover, mice in which telomere length has been artificially reduced exhibit features of aging, as in patients with DC.13,14,38 However, we here demonstrate that telomere dysfunctions in patients with HH are heterogeneous and not systematically associated with critically short telomeres. Indeed, unlike HH3 and the TIN2-deficient patient HH4, patients HH1, HH2, and HH5 exhibit telomere

lengths comparable with those seen in control subjects and in the normal range both in terms of fibroblasts (although presenting with dysfunctional telomere phenotype) and whole blood cells. The association of severe forms of DC caused by telomere dysfunction with normal telomere length in whole blood cells is, to our knowledge, unprecedented. These observations led us to propose a model in which the common characteristics of DC and HH

480 TOUZOT ET AL

J ALLERGY CLIN IMMUNOL FEBRUARY 2012

FIG 5. Telomere length of whole blood cells from patients with HH. A, The mean telomere length of whole blood cells from patients with HH (HH1, HH2, HH3, HH4, and HH5), 15 pediatric control subjects (Ctl1-Ctl15, with ages ranging from 3 months to 9 years), and a TIN2-deficient control subject (TIN2 1/2; age 3 years) was estimated by using the TRF method. The estimated mean telomere length is indicated. B, Graphic representation of TRF data. Green dots represent the TRF values from patients with HH (HH1, HH2, and HH5) comparable with those found in control subjects. Red dots represent the TRF values from patients with HH (HH3 and HH4) with critically short telomeres (<5 kb).

could be the early onset of cellular senescence induced either by critically short telomeres or damaged/unprotected telomeres of otherwise normal length (Fig 6). This assumption implies that telomere length measurements from blood cells commonly used to diagnose DC/HH, although informative, might not identify a subset of patients with HH. Of note, a subset of patients with mutations in C16orf57 and presenting some DC features, which are less severe than those observed in patients with HH, have been recently reported to have normal telomere length.40 Notably, however, telomere dysfunction has not been evaluated in C16orf57-deficient patients. Consequently, there is no evidence that the syndrome caused by C16orf57 mutations is indeed linked to a telomere defect, unlike HH. Blood cells from TIN2-deficient patients exhibit excessive telomere attrition.27,32 We extend this observation to primary fibroblasts from a TIN2-deficient patient (patient HH4). This further argues that DC/HH diseases caused by a bona fide telomere length defect, as is the case for Dyskerin, TERC, TERT, TCAB1, or TIN2 deficiencies,25,29 leads to accelerated telomere shortening not only in blood cells but also in fibroblasts. The telomeric replication is threatening for the genome stability in particular because the replication fork progression at the telomere is unidirectional.41 Consequently, no converging replication forks would be available to help relieve the block when the fork has stalled. Hence several DNA repair factors have been shown to play a key role in telomere protection/replication.1 However, in our assays no defect in DNA DSBs or DNA ICLs repair was detected in cells from patients with HH, ruling out an impairment of these general DNA repair pathways. Nevertheless, we

noted a significant increase in 53BP1 foci in untreated fibroblasts from all but 1 patient with HH, likely reflecting a DNA damage accumulation at the telomere (TIFs), as well as elsewhere in the genome, as recently reported in fibroblasts from Dyskerin-deficient patients as well.42 This is reminiscent of the cellular phenotype described in mice in which the DNA repair or telomere maintenance is compromised, as well as cells from aged wild-type animals.43,44 Because the structural properties of the telomeric G-rich and Crich strands are distinct, the leading- and lagging-strand telomeres of mammalian cells behave differently.5 The DNA replication processes used by the leading and lagging mechanisms also differ.45 CO-FISH experiments reveal the telomeric aberrations specifically affecting one of these 2 strands, as exemplified in cells defective in the Werner helicase or in the flap endonuclease 1 endonuclease in which a specific loss of lagging strand–replicating sister telomeres (C-rich strand) has been observed.46,47 In contrast, TRF2 or Apollo depletion led to fusion between 2 leading-strand telomeres (G-rich strand).37,48 Our CO-FISH study revealed further heterogeneity among patients with HH. Cells from patient HH2 exhibited an increase in telomere doublets predominantly affecting the telomere replicated by leading-strand DNA synthesis. Telomere doublets have been proposed as a signature of impaired telomeric replication, although the underlying molecular process responsible for it is still unknown. Cells from patient HH5 exhibited yet undescribed interstitial telomeric sequences and multiple telomeric signal indicating a profound telomeric instability, a phenotype recently defined in TRF1deficient mice as ‘‘fragile telomeres.’’39,41 However, the TRF1encoding gene was sequenced in cells from patient HH5, and

TOUZOT ET AL 481

J ALLERGY CLIN IMMUNOL VOLUME 129, NUMBER 2

FIG 6. Model proposing the different molecular origins leading to the induction of a DC/HH cellular phenotype. The TIFs and senescence are indicative of telomere dysfunction and are the common features of cells from patients with DC/HH. Telomere dysfunction can originate from a defect of a factor involved in telomere length regulation, rendering the telomere uncapped when reaching too short a length, as observed in cells from patient HH3, cells from patient HH4, and Dyskerin-deficient, TERT-deficient, TCAB1deficient, and TERC-deficient cells (upper panel). Telomere dysfunction can also result from a defect in a telomere-capping factor that leads to DNA damage accumulation at the telomere without accelerated telomere attrition, as observed in patients HH1, HH2, and HH3 (lower panel). Whatever the cause of cellular telomere dysfunction, this could lead to the DC and HH phenotype. Cancer development could ultimately occur when P53-dependent senescence and/or apoptosis pathways become deficient.

found to be normal. The telomeric aberrations seen in cells from patients HH2 and HH5 affect predominantly the leading strand, suggesting a preponderant defect of the leading DNA (G-rich strand) synthesis during telomeric replication. In conclusion, our study provides evidence that severe forms of DC/HH can result from heterogeneous telomere defects not inevitably associated with critical telomere shortening. indicating that the telomere length measurement in blood cells cannot be used as the sole diagnostic tool for this disease. In addition, our results imply that additional factors, some of which are likely involved in telomere protection/replication, are responsible for other forms of HH. The identification of the molecular defects causing other HH conditions in the future should help us understand further the physiology of telomeres and the cause of DC and HH and help us improve their therapy. Clinical implications: Telomere length measurements from blood cells, which are commonly used to diagnose DC/HH caused by telomere dysfunction, although informative, might not identify a subset of patients with DC/HH.

REFERENCES 1. de Lange T. How telomeres solve the end-protection problem. Science 2009;326: 948-52. 2. Blackburn EH. Switching and signaling at the telomere. Cell 2001;106:661-73. 3. Collado M, Blasco MA, Serrano M. Cellular senescence in cancer and aging. Cell 2007;130:223-33. 4. de Lange T. Shelterin: the protein complex that shapes and safeguards human telomeres. Genes Dev 2005;19:2100-10. 5. Gilson E, Geli V. How telomeres are replicated. Nat Rev Mol Cell Biol 2007;8: 825-38. 6. Lansdorp PM. Telomeres and disease. EMBO J 2009;28:2532-40. 7. Walne AJ, Dokal I. Dyskeratosis congenita: a historical perspective. Mech Ageing Dev 2008;129:48-59.

8. Alter BP, Giri N, Savage SA, Rosenberg PS. Cancer in dyskeratosis congenita. Blood 2009;113:6549-57. 9. Revy P, Busslinger M, Tashiro K, Arenzana F, Pillet P, Fischer A, et al. A syndrome involving intrauterine growth retardation, microcephaly, cerebellar hypoplasia, B lymphocyte deficiency, and progressive pancytopenia. Pediatrics 2000;105: E39. 10. Kirwan M, Dokal I. Dyskeratosis congenita, stem cells and telomeres. Biochim Biophys Acta 2009;1792:371-9. 11. Zhong F, Savage SA, Shkreli M, Giri N, Jessop L, Myers T, et al. Disruption of telomerase trafficking by TCAB1 mutation causes dyskeratosis congenita. Genes Dev 2010;25:11-6. 12. Hao LY, Armanios M, Strong MA, Karim B, Feldser DM, Huso D, et al. Short telomeres, even in the presence of telomerase, limit tissue renewal capacity. Cell 2005;123:1121-31. 13. Blasco MA. Mice with bad ends: mouse models for the study of telomeres and telomerase in cancer and aging. EMBO J 2005;24:1095-103. 14. Armanios M, Alder JK, Parry EM, Karim B, Strong MA, Greider CW. Short telomeres are sufficient to cause the degenerative defects associated with aging. Am J Hum Genet 2009;85:823-32. 15. Atzmon G, Cho M, Cawthon RM, Budagov T, Katz M, Yang X, et al. Evolution in health and medicine Sackler colloquium: genetic variation in human telomerase is associated with telomere length in Ashkenazi centenarians. Proc Natl Acad Sci U S A 2009;107(suppl 1):1710-7. 16. Tomas-Loba A, Flores I, Fernandez-Marcos PJ, Cayuela ML, Maraver A, Tejera A, et al. Telomerase reverse transcriptase delays aging in cancer-resistant mice. Cell 2008;135:609-22. 17. Touzot F, Callebaut I, Soulier J, Gaillard L, Azerrad C, Durandy A, et al. Function of Apollo (SNM1B) at telomere highlighted by a splice variant identified in a patient with Hoyeraal-Hreidarsson syndrome. Proc Natl Acad Sci U S A 2010;107: 10097-102. 18. Buck D, Malivert L, de Chasseval R, Barraud A, Fondaneche MC, Sanal O, et al. Cernunnos, a novel nonhomologous end-joining factor, is mutated in human immunodeficiency with microcephaly. Cell 2006;124:287-99. 19. Hoyeraal HM, Lamvik J, Moe PJ. Congenital hypoplastic thrombocytopenia and cerebral malformations in two brothers. Acta Paediatr Scand 1970;59:185-91. 20. Hreidarsson S, Kristjansson K, Johannesson G, Johannsson JH. A syndrome of progressive pancytopenia with microcephaly, cerebellar hypoplasia and growth failure. Acta Paediatr Scand 1988;77:773-5. 21. Ohga S, Kai T, Honda K, Nakayama H, Inamitsu T, Ueda K. What are the essential symptoms in the Hoyeraal-Hreidarsson syndrome? Eur J Pediatr 1997;156:80-1.

482 TOUZOT ET AL

22. Pinto FO, Leblanc T, Chamousset D, Le Roux G, Brethon B, Cassinat B, et al. Diagnosis of Fanconi anemia in patients with bone marrow failure. Haematologica 2009;94:487-95. 23. Arnoult N, Shin-Ya K, Londono-Vallejo JA. Studying telomere replication by Q-CO-FISH: the effect of telomestatin, a potent G-quadruplex ligand. Cytogenet Genome Res 2008;122:229-36. 24. Goldman F, Bouarich R, Kulkarni S, Freeman S, Du HY, Harrington L, et al. The effect of TERC haploinsufficiency on the inheritance of telomere length. Proc Natl Acad Sci U S A 2005;102:17119-24. 25. Armanios M, Chen JL, Chang YP, Brodsky RA, Hawkins A, Griffin CA, et al. Haploinsufficiency of telomerase reverse transcriptase leads to anticipation in autosomal dominant dyskeratosis congenita. Proc Natl Acad Sci U S A 2005;102: 15960-4. 26. Vulliamy T, Marrone A, Szydlo R, Walne A, Mason PJ, Dokal I. Disease anticipation is associated with progressive telomere shortening in families with dyskeratosis congenita due to mutations in TERC. Nat Genet 2004;36:447-9. 27. Walne AJ, Vulliamy TJ, Beswick R, Kirwan M, Dokal I. TINF2 mutations result in very short telomeres: analysis of a large cohort of patients with dyskeratosis congenita and related bone marrow failure syndromes. Blood 2008;112:3594-600. 28. Takai H, Smogorzewska A, de Lange T. DNA damage foci at dysfunctional telomeres. Curr Biol 2003;13:1549-56. 29. Mitchell JR, Wood E, Collins K. A telomerase component is defective in the human disease dyskeratosis congenita. Nature 1999;402:551-5. 30. Kim NW, Wu F. Advances in quantification and characterization of telomerase activity by the telomeric repeat amplification protocol (TRAP). Nucleic Acids Res 1997;25:2595-7. 31. d’Adda di Fagagna F, Reaper PM, Clay-Farrace L, Fiegler H, Carr P, Von Zglinicki T, et al. A DNA damage checkpoint response in telomere-initiated senescence. Nature 2003;426:194-8. 32. Savage SA, Giri N, Baerlocher GM, Orr N, Lansdorp PM, Alter BP. TINF2, a component of the shelterin telomere protection complex, is mutated in dyskeratosis congenita. Am J Hum Genet 2008;82:501-9. 33. Allsopp RC, Vaziri H, Patterson C, Goldstein S, Younglai EV, Futcher AB, et al. Telomere length predicts replicative capacity of human fibroblasts. Proc Natl Acad Sci U S A 1992;89:10114-8. 34. O’Sullivan RJ, Karlseder J. Telomeres: protecting chromosomes against genome instability. Nat Rev Mol Cell Biol 2010;11:171-81.

J ALLERGY CLIN IMMUNOL FEBRUARY 2012

35. Palm W, de Lange T. How shelterin protects mammalian telomeres. Annu Rev Genet 2008;42:301-34. 36. Revy P, Buck D, le Deist F, de Villartay JP. The repair of DNA damages/modifications during the maturation of the immune system: lessons from human primary immunodeficiency disorders and animal models. Adv Immunol 2005;87:237-95. 37. Bailey SM, Cornforth MN, Kurimasa A, Chen DJ, Goodwin EH. Strand-specific postreplicative processing of mammalian telomeres. Science 2001;293:2462-5. 38. Blasco MA, Lee HW, Rizen M, Hanahan D, DePinho R, Greider CW. Mouse models for the study of telomerase. Ciba Found Symp 1997;211:160-76. 39. Martinez P, Thanasoula M, Munoz P, Liao C, Tejera A, McNees C, et al. Increased telomere fragility and fusions resulting from TRF1 deficiency lead to degenerative pathologies and increased cancer in mice. Genes Dev 2009;23:2060-75. 40. Walne AJ, Vulliamy T, Beswick R, Kirwan M, Dokal I. Mutations in C16orf57 and normal-length telomeres unify a subset of patients with dyskeratosis congenita, poikiloderma with neutropenia and Rothmund-Thomson syndrome. Hum Mol Genet 2010;19:4453-61. 41. Sfeir A, Kosiyatrakul ST, Hockemeyer D, MacRae SL, Karlseder J, Schildkraut CL, et al. Mammalian telomeres resemble fragile sites and require TRF1 for efficient replication. Cell 2009;138:90-103. 42. Kirwan M, Beswick R, Walne AJ, Hossain U, Casimir C, Vulliamy T, et al. Dyskeratosis congenita and the DNA damage response. Br J Haematol 2011;153: 634-43. 43. Nijnik A, Woodbine L, Marchetti C, Dawson S, Lambe T, Liu C, et al. DNA repair is limiting for haematopoietic stem cells during ageing. Nature 2007;447:686-90. 44. Rossi DJ, Bryder D, Seita J, Nussenzweig A, Hoeijmakers J, Weissman IL. Deficiencies in DNA damage repair limit the function of haematopoietic stem cells with age. Nature 2007;447:725-9. 45. Moser BA, Subramanian L, Chang YT, Noguchi C, Noguchi E, Nakamura TM. Differential arrival of leading and lagging strand DNA polymerases at fission yeast telomeres. EMBO J 2009;28:810-20. 46. Saharia A, Guittat L, Crocker S, Lim A, Steffen M, Kulkarni S, et al. Flap endonuclease 1 contributes to telomere stability. Curr Biol 2008;18:496-500. 47. Crabbe L, Verdun RE, Haggblom CI, Karlseder J. Defective telomere lagging strand synthesis in cells lacking WRN helicase activity. Science 2004;306: 1951-3. 48. Wu P, van Overbeek M, Rooney S, de Lange T. Apollo contributes to G overhang maintenance and protects leading-end telomeres. Mol Cell 2010;39:606-17.

J ALLERGY CLIN IMMUNOL VOLUME 129, NUMBER 2

TOUZOT ET AL 482.e1

FIG E1. A, Protein sequence alignment of the wild-type human TIN2 (upper sequence) and the mutated TIN2 identified in patient HH4 (bottom sequence). The TRF1-binding motif (TBM; gray box) is indicated. B, Scheme representing wild-type and truncated TIN2 resulting from the c838 A>T (K280X) mutation.

482.e2 TOUZOT ET AL

J ALLERGY CLIN IMMUNOL FEBRUARY 2012

FIG E2. Representative images of senescence and TIFs, 2 markers of telomere dysfunction. Telomere dysfunction leads to senescence as detected by the senescence-associated b-galactosidase activity on the right panel (primary fibroblasts from HH2; A) and TIFs corresponding to the colocalization of 53BP1 and TRF2 labeling (yellow arrows on bottom panels; B). Fibroblasts from a healthy donor (5 years) were analyzed at passage number 9.

J ALLERGY CLIN IMMUNOL VOLUME 129, NUMBER 2

TOUZOT ET AL 482.e3

FIG E3. Representative images of 53BP1 foci detected by means of immunofluorescence in fibroblasts without treatment, 1 hour, and 24 hours after 5 Gy x-ray irradiation. A significant number of 53BP1 foci were still observed in Cernunnos-deficient fibroblasts but not in cells from control subjects and patients with HH.