Abstract
purpose. To investigate whether individuals, with no family history of ataxia telangiectasia (AT), in whom idiopathic or radiation-induced ocular telangiectasia developed are carriers of ATM gene mutations.
methods. The ATM cDNA from lymphoblastoid cell lines established from 16 patients with idiopathic retinal or choroidal telangiectasia and 14 patients with radiation-induced telangiectasia after radiotherapy for age-related macular degeneration (AMD) was screened using the restriction endonuclease fingerprinting technique. The frequency of each detected variant was determined in the French population by either a mass spectrometry-based technique or variant-specific endonuclease digestion.
results. Twenty-one ATM missense alterations, at 10 different sites, 8 of which would result in an amino acid substitution at a conserved position in the ATM protein were found. Four were novel changes, three of which were not detected in the 128 French control subjects screened. Eleven of 16 of the individuals with either idiopathic polypoidal choroidal vasculopathy or juxtafoveolar retinal telangiectasis and 6 of 14 individuals that had choroidal telangiectasis after radiotherapy for AMD carried ATM sequence variants. These latter six individuals had a significantly shorter delay time before the presentation of this vasculopathy compared with those individuals who had a wild-type ATM (11.8 ± 3.4 months vs. 17.5 ± 4.5 months, P = 0.024). They had also received a lower average dose of X-rays, although this difference did not reach statistical significance (18.7 ± 3.9 Gy vs. 23.7 ± 5.6 Gy, P = 0.09).
conclusions. ATM missense variants could confer an AT-like phenotype and influence the formation of retinal and choroidal vascular abnormalities.
One of the clinical characteristics of the radiation-sensitive, cancer-prone disease ataxia telangiectasia (AT), caused in the majority of cases by mutations in the
ATM gene (
ATM; Mendelian Inheritance in Man [MIM] 208900; http://www.ncbi.nlm.nih.gov/Genbank; provided in the public domain by the National Center for Biotechnology Information, Bethesda, MD), is the development of telangiectases in the eyes and ears.
1 This disease is extremely rare, with the incidence of AT being estimated to be in the range between 1 per 40,000 and 1 per 300,000 live births in various populations.
2 3 However, the frequency of
ATM heterozygotes in the general population, based on both epidemiologic studies and more recently on direct mutation analysis of the
ATM gene in control populations, is of the order of 1%. Although heterozygous carriers of the
ATM gene show none of the clinical features of AT, there has been consistent evidence since the first report by Swift et al
. 4 that they have a higher risk of development of cancer, particularly of the female breast (reviewed in Angèle and Hall
5 ).
The
ATM gene has 66 exons scanning 150 kb of genomic DNA on chromosome 11 and is expressed in a wide range of tissues as an approximately 13-kb messenger RNA encoding a 350-kDa protein serine/threonine protein kinase.
6 ATM’s kinase activity is itself enhanced in response to DNA double-strand breaks, resulting in a phosphorylation cascade activating many proteins, each of which in turn affects a specific signaling pathway. These substrates include the protein products of a number of well-characterized tumor-suppressor genes including
TP53,
BRCA1, and
CHK2, which play important roles in triggering cell cycle arrest, DNA repair, or apoptosis (reviewed by Shiloh and Kastan
7 ). At the cellular level, cell lines established from patients with AT show hypersensitivity to ionizing radiation (IR) and other agents that cause DNA double-strand breaks with defects in both G
1/S and G
2/M checkpoint responses, defective inhibition of DNA replication and increased chromosomal instability being seen after such exposures (reviewed in Shiloh and Kastan
7 and Khanna et al.
8 ).
The profile of
ATM mutations in children with AT has revealed that most are unique and uniformly distributed along the length of the gene. Most patients with AT are compound heterozygotes with two different
ATM mutations; patients homozygous for the same
ATM mutation are rare.
7 The predominate type of mutation found in the
ATM gene in patients with AT results in a truncated and unstable ATM protein. However, a considerable percentage of
ATM alleles from patients with AT contain in-frame insertions/deletions or missense mutations. It would be expected that many such changes would give rise to the expression of a stable, albeit mutant, form of the ATM protein. The presence of such a mutant protein with residual function, or the presence of a small amount of normal protein, has been found in some cases of AT to lead to a milder clinical presentation and altered cancer predisposition.
9
ATM heterozygotes in the general population thus fall into two groups distinguished by their mutation profile.
10 11 One group is heterozygous for a truncating allele, which would be expected to give rise to little or no protein and could lead to ATM haploinsufficiency.
12 The second group is heterozygous for a missense allele, resulting in amino acid substitutions or short in-frame deletions or insertions and the expression of some ATM protein that may act in a dominant-negative manner competing with the wild-type ATM protein. It has been predicted that carriers of such an allele would have an increased cancer risk, as has been seen for the 7271G→T missense
ATM sequence alteration that is associated with an increased breast cancer risk
9 13 or perhaps of developing clinical phenotypes associated with AT, such as radiation sensitivity and telangiectases or vasculopathies. To test this hypothesis we performed a pilot study in which we investigated whether 30 individuals with no family history of AT, in whom idiopathic or radiation-induced ocular telangiectasia developed of either choroidal or retinal origin, were carriers of
ATM mutations.
The protocol of this study was reviewed and approved by the local ethics committees, and informed consent was obtained from all the study participants. This study adhered to the tenets of the Declaration of Helsinki. The 30 individuals (9 males and 21 females) included in the pilot study had either choroidal or retinal telangiectasia
(Table 1) . Within the group with choroidal telangiectasia was one subset of eight patients who had idiopathic polypoidal choroidal vasculopathy (IPCV). These individuals were diagnosed by biomicroscopic fundus examination and fluorescein and indocyanine green (ICG) angiographies.
The second subset (14 patients) in this group had age-related macular degeneration (AMD) with subfoveal choroidal neovascularization as a complication. They were treated with external beam radiotherapy and had subsequent development of radiation-induced choroidal telangiectasia (RICT).
14 15 RICT was diagnosed by microscopic fundus examination and ICG angiography. These complications were detected relatively late after radiotherapy, with the average onset in the group in the present study being 15.1 ± 4.9 months after radiotherapy.
The second group of eight patients included in this study had idiopathic juxtafoveolar retinal telangiectasia (IJRT), diagnosed by biomicroscopic fundus examination and fluorescein angiography. In two patients, the IJRT was unilateral and in six bilateral and in all individuals resulted in the formation of macular edema.
Lymphocytes were separated from fresh blood samples with a single-density gradient (Ficoll; Pharmacia, Uppsala, Sweden) and a lymphoblastoid cell line (TEL line) established by infection with Epstein-Barr virus in each patient. The cell lines were routinely cultured in RPMI 1640 medium supplemented with 10% fetal calf serum (FCS) in an atmosphere of 5% CO2.
To determine the frequency of specific ATM sequence changes in the French population, genomic DNA was isolated from the buffy coat of blood samples obtained from 49 male and 80 female blood donors living in the Rhône area of France by using separation columns (Qiagen SA, Courtaboeuf, France) and following the manufacturer’s protocol.
RT-PCR of the ATM Open Reading Frame.
Total RNA was extracted from each TEL line (TRIzol; Invitrogen, SARL, Cergy Pontoise, France). The RT-PCR was performed with reverse transcriptase (Superscript II; Invitrogen) according to the manufacturer’s recommendations. The reaction products were used as a template to amplify the entire ATM open reading frame (ORF) using the following primers, ATM A: 5′-atgagtctagtacttaatgatctgc-3′; and ATM B: 5′-aaggctgaatgaaagggtaattc-3′. The PCR reaction was performed in 25 μL containing 1 μL of the RT product, 1× PCR buffer, 1 μM of each primer, 400 μM of each dNTP, 2 U of a polymerase (Expand Long Template; Roche Diagnostics, Meylan, France) and 0.66 μg of a polymerase antibody (TaqStart; Clonetech-BD Biosciences, Bedford, MA).
Restriction Endonuclease Fingerprinting Analysis.
The diluted RT-PCR product of 9205 bp was used as a template for the amplifications, by nested PCR, of eight overlapping fragments of 1.0 to 1.6 kb that were analyzed by the restriction endonuclease fingerprinting (REF) technique.
16 The primers used for the eight nested PCRs were as previously described by Gilad et al.
17 except for primers 1a, 1b, 6b, and 7a which were: 1a, ATM A; 1b, 5′-ggtctgcaggctgacccagtaaataac-3′; 6b, 5′-gataggtctgcatgatgaccg-3′; and 7a, 5′-aatgatcaagaagttggatgcc-3′. The PCR product for each fragment was digested in five separate tubes with a pool of one to three different restriction endonucleases (list of enzymes available on request) overnight at 37°C in the presence of 0.2 U shrimp alkaline phosphatase (Roche Diagnostics). The individual digests were pooled and labeled using 5 μCi of γ-[
33P] adenosine triphosphate (ATP) and 5 U of T4 polynucleotide kinase (New England Biolabs, Inc., Beverly, MA) for 45 minutes at 37°C. The labeled products were denatured for 3 minutes at 90°C and then analyzed under three electrophoresis conditions: 0.5× and 0.8× mutation-detection electrophoresis (MDE) gels without glycerol and a 0.8× MDE gel with 10% glycerol, in 0.6× TBE (5.34 mM Tris, 0.15 mM EDTA, and 5.34 mM boric acid). The electrophoresis was performed at a constant power of 20 W for 4 hours for the 0.5× MDE gel, 20 W for 6 hours for the 0.8× MDE gel without glycerol, and 14 W overnight for the 0.8× MDE gel containing 10% glycerol at 22° to 24°C. Gels were dried and subjected to autoradiography for 12 hours at room temperature.
All fragments with an altered REF pattern were sequenced using a ready reaction cycle sequencing kit (Prism Big Dye Terminator ver. 2.0; Applied Biosystems, Courtaboeuf, France) to determine the exact base changes. Primers used for the amplification of the different exons were those described by Sandoval et al.
18
Determination of the Frequencies of ATM Sequence Changes in Control subjects.
The frequency of each
ATM sequence change found in the TEL cell lines was assessed in a French control group, using either a mass spectrometry-based technique, short oligonucleotide mass analysis (SOMA),
19 or endonuclease digestion after PCR-mediated, site-directed mutagenesis (PSDM).
20
SOMA is a specific and sensitive genotyping method that allows the simultaneous analysis of three polymorphisms using multiplex-PCR and electrospray ionization tandem mass spectrometry (ESI-MS/MS) analysis.
19 Genomic DNA (40 ng) was amplified with primers
(Table 2) containing the recognition site (CTGGAG) for the type-IIs endonuclease
BpmI in a reaction volume of 25 μL containing 2 mM MgCl
2, all four deoxyribonucleoside triphosphates each at 200 μM, 0.5 μM of each primer, and 2.5 U of
Taq polymerase (Platinum
Taq; Invitrogen). Cycling conditions were 94°C for 2 minutes followed by 40 cycles of 94°C for 30 seconds, 60°C for 30 seconds, and 72°C for 30 seconds, with a final extension at 72°C for 5 minutes. The PCR product was digested overnight at 37°C with 5 U
BpmI (New England Biolabs, Inc.) or at 30°C with 5 U of its isoenzyme
GsuI (Fermentas Inc., Hanover, MD). The 7-, 8-, or 9-mer oligonucleotides generated
(Table 2) were then characterized by ESI-MS/MS based on both the molecular weight and sequence.
Before analysis by ESI-MS/MS, the pellet was resuspended in 6 μL high-performance liquid chromatography (HPLC) mobile phase (80:20 vol/vol; solvent A to solvent B), where solvent A was 0.4 M 1,1,1,3,3,3-hexafluoro-2-propanol (HFIP, pH 7.0) (Roth-Sochiel, Lauterberg, France) and solvent B was 0.8 M HFIP with methanol (50:50 vol/vol). Samples (5 μL) were introduced into the mass spectrometer through HPLC with a 15 cm × 800 μm microcolumn (Vydac C18; LC Packings, Amsterdam, The Netherlands) operated at a flow rate of 20 μL/min. The HPLC gradient was increased from an initial 20% B to 75% B at 7.1 minutes. After a 2-minute holding phase, the gradient was returned to initial conditions. Oligonucleotide peaks eluted at approximately 8 minutes.
Mass spectra were obtained on a triple-quadruple mass spectrometer (Quattro-LC; Micromass, Manchester, UK) equipped with an electrospray ion source operated in the negative ion mode. The capillary voltage was held at 5 kV and the cone voltage at 77 volts. Argon pressure in the collision cell was 4.2 × 10
−3 mBar, and collision energy was held at 38 volts. The mass spectrometer was set to less than unit resolution. Data were collected by selected-reaction monitoring, using transitions for each oligonucleotide as shown in
Table 2 .
PSDM allows the presence of an altered allele to be detected by the ability of a restriction endonuclease to cut a PCR fragment containing a defined polymorphism. When the polymorphism does not create or abolish a restriction enzyme recognition site, primers can be designed that create an allele-specific restriction enzyme recognition site.
20 This approach was used to determine the frequency of the
ATM sequence variants A2512G and C4578T. The primers and the corresponding restriction enzyme are listed in
Table 2 . The lowercase g and c are mismatches that were specifically introduced to create
ClaI and
StyI restriction sites, respectively. In both cases, only the wild-type allele is cut by the endonuclease. The PCR product was digested with
ClaI and
StyI, according to the manufacturer’s instructions, and the fragments analyzed by electrophoresis on a 3% agarose gel (NuSieve; BioWittaker, Walkersville, MD). DNA samples known to be carrying the mutant allele were included in each analysis, with the genotype of the control samples being determined by the banding pattern observed on the gels. A random sample of DNAs were also analyzed by direct sequencing of the corresponding exon and complete concordance between the different techniques was observed (data not presented). The frequency of each sequence change was determined in at least 120 healthy blood donors.
The efficiency of the REF mutation screening technique has been estimated to be on the order of 70%, depending on the electrophoresis conditions used.
21 In the present study, each REF fragment was analyzed under three different electrophoresis conditions, thus maximizing mutation detection. Using such an approach, we found 21
ATM missense sequence variants, at 10 different sites on the
ATM gene, in 17 of 30 TEL lines examined. Eight of 10 of the sequence changes would be predicted to result in an amino acid substitution in the ATM protein at a conserved position, whereas the 735C→T and 4578C→T variants were silent alterations
(Tables 1 3) . No truncating
ATM mutations were found.
Four of the sequence changes in the TEL lines were novel
(Table 3) —that is, not reported in the latest update of the
ATM mutation database (http://www.benaroyaresearch.org/bri_investigators/atm.htm hosted by the Benaroya Research Institute at Virginia Mason, Seattle, WA). Three of the novel changes: A2512G, G5262T, and G6315C were not detected in the French control subjects screened and thus may represent mutations or rare polymorphisms. The genotype distribution of all the polymorphisms was consistent with Hardy-Weinberg equilibrium.
The most frequently found alteration was the 5557G→A polymorphism, which was present in 9 of 30 of the TEL lines (30%). However, there were no significant differences in the frequency of the A allele between the TEL patients (A allele frequency 0.167) and the population control subjects (A allele frequency 0.128). Three of the TEL patients carried more than one ATM sequence variant, always in combination with the 5557A allele. The 998T allele was detected only once in the TEL patients. This same genotype combination (998T, 5557A) was also found in the only two control individuals carrying this rare sequence variant. The 4578T allele was present in combination with the 5557A allele in one patient, a combination that has not been found in the 13 carriers of the 4578T allele in the control group. The third patient carried multiple changes: the novel variant 2512G was detected together with the 4258T and 5557A alleles. Neither the 2512G allele nor the combination of 4258T and 5557A was detected in the control population.
Six of 14 patients in whom RICT developed after radiotherapy for AMD had ATM sequence variants, five of which would change the coding potential of the ATM gene. These six individuals showed a significantly shorter delay time before the presentation of this vasculopathy, compared with those who had wild-type ATM (11.8 ± 3.4 months vs. 17.5 ± 4.5 months, P = 0.026). They had also received a lower average dose of X-rays than the group with wild-type ATM, although this difference did not reach statistical significance (18.7 ± 3.4 Gy vs. 23.7 ± 5.7 Gy, P = 0.09).
The purpose of this pilot study was to investigate whether individuals, with no family history of AT, who had idiopathic or radiation-induced ocular telangiectasia of either choroidal or retinal origin, were carriers of
ATM mutations. The molecular origins of these telangiectases are poorly understood. Three case reports have described the clinical features of siblings affected by bilateral idiopathic juxtafoveolar retinal telangiectasia (IJFT),
22 23 24 suggesting a genetic influence on the pathogenesis of this disease. The main systemic association noted for IJFT, reported in two case series, is with diabetes.
25 26 One of the few identified risk factors for the development of IJFT is previous radiation exposure.
27
IPCV is a chronic disease with a diverse clinical spectrum and a variable outcome.
28 29 It can be associated with extensive alterations of the retinal pigment epithelium as seen in individuals with diffuse retinal epitheliopathy, or with the presence of drusen-hyaline excrescences found in the Bruch’s membrane as seen in patients with AMD.
30 31 These polypoidal vascular dilations are located in the postcapillary choroidal veins. A recent histopathology study showed that these polypoidal structures are composed of saccular dilated thin-walled blood vessels without pericytes. The positive immunohistochemical staining for vascular endothelial growth factor in the retinal pigment epithelium and the vascular endothelial cells suggest that this vascular complex is a subretinal choroidal neovascularization.
32
Telangiectasic dilations of the vessels in the outer portions of the choroidal neovascular network has been reported as a complication in patients who have undergone radiotherapy for subfoveal choroidal neovascularization secondary to AMD.
14 15 33 In the cohort of patients observed by Spaide et al.,
33 who were treated either in New York, where most patients were treated with 10 Gy, 12.6% of patients exhibited development of a vascular abnormality, or in Belgium, where the patients were treated with 20 Gy, 7.1% had abnormalities. The appearance of these abnormalities was 3 to 9 months after external beam radiation in New York and detailed as somewhat later in the Belgian group of patients.
33 The patients with RICT included in this present study had been treated with doses (between 15 and 28.8 Gy in four to eight fractions) that more closely resembled those used in Belgium, with a total of 19 (6.4%) eyes of the 295 eyes treated developing RICT. These choroidal telangiectases were not present in any of these patients in the first 6 months after irradiation.
14 15 The radiation dose and fractionation schedules thus appear to be determinants for the time of onset, rate of progression, and severity of these vasculopathies. It has been suggested that concomitant chemotherapy and preexisting diabetes may exaggerate the vasculopathy by intensifying the oxygen-derived free-radical assault on the vascular cells.
34
The most frequent sequence variant detected in the
ATM gene in the study population was the common polymorphism 5557G→A. The 5557A allele has been found to modulate the penetrance of hereditary nonpolyposis colorectal cancer in carriers of germline
MLH1 and
MSH2 mutations.
35 It has also been detected more frequently in patients with prostate cancer who had a severe late responses to radiation therapy: 35% of the radiosensitive cases compared with 18% in those who received comparable radiotherapy but who did not have late sequelae.
36 It has been suggested that the 3161G variant allele found in one TEL patient, may be more common in breast cancer cases selected for first-degree family history and early age of onset.
37 However a recent study by Spurdle et al.
38 found no evidence for an association of this variant with a moderate or high risk of breast cancer in Australian women. It remains to be established whether these nonconservative amino acid substitution variants result in altered ATM function.
The frequency of all the
ATM sequence variants identified in the present study population was determined in the French general population and, for three variants, no carriers were identified in more than 120 healthy control subjects, suggesting that these are
ATM mutations or rare polymorphisms. Unfortunately, the power of the study is insufficient to be able to detect differences between the patients and the control subjects when the prevalence of the rare allele is less than 6%. Although the number of individuals examined in each group was small, 9 (56%) of 16 of the patients with idiopathic retinal or choroidal telangiectasia had sequence alterations that would be predicted to change the amino acid sequence of the ATM protein. This frequency is higher than that recently reported in by Dork et al.,
39 who found that 46% of the 192 breast cancer patients screened carried
ATM missense variants. Three individuals with idiopathic telangiectasia also had an exonic sequence that would not change the
ATM coding sequence. However the 735C→T substitution has been suggested to enhance the alternative splicing of exon 9
40 and was found 5.6 times more frequently in individuals with diffuse large B-cell lymphoma than in individuals in random cases.
41
In summary, a high frequency of carriers of rare
ATM missense sequence variants was found within these groups of patients in whom ocular telangiectasia of either choroidal or retinal origin developed. Spaide et al.
33 have suggested that the RICT arises as a growth of new blood vessels at the border of the older area of choroidal neovascularization. Radiation has been shown to affect vascular endothelial cells in vitro inducing morphologic responses, DNA breaks, and vascular endothelial cell apoptosis.
33 Thus, although the radiotherapy may temporarily affect the ability of the blood vessels to grow, the neovascularization stimuli probably remain, resulting in the eventual reactivation of vessel growth causing the RICT that was present in the patients. The mode and action of these stimuli remain to be established but may be related to alterations in the cellular responses to oxidative stress, as has been noted in patients with AT in which ocular telangiectases are one of the characteristic clinical manifestations.
42 There is mounting evidence for oxidative stress in ATM-deficient cells, although the mechanism underlying ATM’s involvement and the interplay with different signaling pathways remains to be established.
7 For instance, a functional link has been established between ATM and the insulin response pathway,
43 and, as discussed earlier, one of the few identified risk factors for the development of IJFT is previous radiation exposure with the main systemic association being with diabetes.
The findings of a high frequency of ATM missense sequence alterations and no truncating mutations in these patients would be consistent with our working hypothesis that missense ATM variants could confer an AT-like phenotype acting in a dominant negative fashion. To our knowledge this is the first report implicating the ATM gene, outside AT-affected families, in a non-neoplastic clinical disorder, and this study has identified a potential susceptibility gene for the development of ocular telangiectasia of either choroidal or retinal origin. A larger study with more cases is planned to assess whether these preliminary findings are significant, with a detailed personal history being taken in particular of past radiation exposure and diabetes. The establishment of lymphoblastoid cell lines from the study group will allow the assessment of whether these rare ATM sequence variants confer altered ATM function.
The authors thank Brigitte Chapot for technical assistance, Paola Pisani, Ruggero Montesano, and Yossi Shiloh for advice, Micheline Absi from the Blood Transfusion Service in Lyon for help in the study.