Mutations in CTC1, Encoding the CTS Telomere Maintenance Complex Component 1, Cause Cerebroretinal Microangiopathy with Calcifications and Cysts

Abstract

Cerebroretinal microangiopathy with calcifications and cysts (CRMCC) is a rare multisystem disorder characterized by extensive intracranial calcifications and cysts, leukoencephalopathy, and retinal vascular abnormalities. Additional features include poor growth, skeletal and hematological abnormalities, and recurrent gastrointestinal bleedings. Autosomal-recessive inheritance has been postulated. The pathogenesis of CRMCC is unknown, but its phenotype has key similarities with Revesz syndrome, which is caused by mutations in TINF2, a gene encoding a member of the telomere protecting shelterin complex. After a whole-exome sequencing approach in four unrelated individuals with CRMCC, we observed four recessively inherited compound heterozygous mutations in CTC1, which encodes the CTS telomere maintenance complex component 1. Sanger sequencing revealed seven more compound heterozygous mutations in eight more unrelated affected individuals. Two individuals who displayed late-onset cerebral findings, a normal fundus appearance, and no systemic findings did not have CTC1 mutations, implying that systemic findings are an important indication for CTC1 sequencing. Of the 11 mutations identified, four were missense, one was nonsense, two resulted in in-frame amino acid deletions, and four were short frameshift-creating deletions. All but two affected individuals were compound heterozygous for a missense mutation and a frameshift or nonsense mutation. No individuals with two frameshift or nonsense mutations were identified, which implies that severe disturbance of CTC1 function from both alleles might not be compatible with survival. Our preliminary functional experiments did not show evidence of severely affected telomere integrity in the affected individuals. Therefore, determining the underlying pathomechanisms associated with deficient CTC1 function will require further studies.

Main Text

Cerebroretinal microangiopathy with calcifications and cysts (CRMCC [MIM 612199]) is a rare multisystem disorder. Its key characteristics include extensive and progressive intracranial calcifications and leukoencephalopathy, and many affected individuals also develop intracranial cysts (Figures 1A–1F).¹ Typically, the cerebral findings are accompanied by retinal vascular abnormalities, which include both retinal telangiectasias—leading to bleedings and exudative retinopathy—and angioma-like excrescences of retinal vessels associated with bleedings and vitreoretinal traction (Figures 1G and 1H).^1,2 The retina that is peripheral to these abnormalities is avascular. Abnormal small vessels with thickened walls and angiomatous proliferations have also been reported in brain areas that show calcifications.¹ Individuals with CRMCC have previously been described as having one of two disorders: Coats plus³ or leukoencephalopathy with calcifications and cysts (LCC).⁴ However, reports in which several individuals were found to have characteristics of both Coats plus and LCC led us to postulate that these two entities are in fact manifestations of the same disorder, CRMCC.^1,5

Figure 1.

Key Clinical Findings in CRMCC

(A) A T2-weighted magnetic resonance image (MRI) of a CRMCC-affected individual (age 11 years) shows heterogeneous lesions in both thalami (black arrows), hyperintensity in the posterior parts of the internal capsules (white arrow) and in the right putamen, and a cyst in the midline (arrowhead).

(B) A T2^∗-weighted image reveals large calcifications in thalami (arrow) and in the parieto-occipital regions.

(C) A T2-weighted image of another individual (age 16 years) with CRMCC shows hyperintense lesions in the brain stem (arrow).

(D and E) Flair images of a 3-year-old individual reveal widespread white-matter hyperintensity in the parieto-occipital regions (arrowheads) as well as asymmetric thalamic involvement (black arrow). A stripe-like hypodensity in the subcortical white matter is seen (white arrow) and most likely represents calcification. Note a midline cyst.

(F) A brain computed tomography (CT) image of a 3-year-old individual shows dense calcifications in the subcortical white matter (arrowheads) and in the brainstem (arrow).

(G and H) Wide-angle fundus photographs of a 15-month-old individual's eyes show a preretinal hemorrhage (arrow) associated with telangiectatic retinal vessels (arrowheads), which border avascular peripheral retina in the right eye (G), and an exudative retinal detachment with yellow lipid exudates in the left eye (H).

(I) Radiographs of the lower limbs of a 14-year-old individual show osteopenia and abnormal bone structure (especially in the metaphyseal areas), bilateral fragility fractures in the proximal tibiae (arrows), and genua valga.

Most CRMCC-affected individuals present in infancy or early childhood with poor growth and neurological symptoms, which include spasticity, dystonia, ataxia, seizures, and obstructive hydrocephalus and/or impaired vision. The neurological symptoms are slowly progressive and, in most individuals, are associated with some degree of cognitive decline.^1,2 Affected children typically show compromised prenatal and postnatal growth. Skeletal abnormalities include generalized osteopenia or early-onset low-turnover osteoporosis, fragility fractures with delayed healing, and metaphyseal abnormalities (Figure 1I).^3,6 Recurrent gastrointestinal hemorrhage due to telangiectatic or angiodysplastic changes is a life-threatening complication that occurs in up to 40% of individuals with CRMCC.^1,2 Additional findings in some individuals are portal hypertension, liver failure, and unexplained anemia and thrombopenia with varying degrees of bone-marrow suppression.^1,2,6 Less often, nail dystrophy, graying, or sparse hair or skin-pigmentation abnormalities have been reported.^2,7,8

CRMCC most often occurs sporadically, but several families with healthy parents and two or more affected children of both sexes have been reported,^1,2,7,8 suggesting autosomal-recessive inheritance. Previously, no genes or variants associated with CRMCC have been reported. The pathogenesis of this pleiotropic disease remains unknown, although obliterative microangiopathy has been suggested as a possible primary underlying pathological abnormality.¹

We employed a whole-exome sequencing strategy in four apparently unrelated Finnish individuals with CRMCC. We then Sanger sequenced 11 additional affected members, three of whom were non-Finnish, from ten families. In addition to the tabulated key features (Table 1), a detailed clinical description of 10 of the 15 subjects can be found in three previous papers.^1,2,9 The study was approved by an institutional review board at the Helsinki University Central Hospital, and written informed consent was obtained from all study subjects or their legal guardians before peripheral blood was drawn for DNA extraction.

Table 1.

Summary of Clinical Features and Mutation Status in 15 CRMCC-Affected Individuals

	F1:II-1	F2:II-1	F3:II-1	F4:II-2a	F5:II-1	F6:II-3	F7:II-3a	F8:II-1	F9:II-2	F10:II-1	F10:II-2	F11:II-1	F12:II-1	F13:II-3	F14: II-1
	Patient 21	Patient 11	Patient 16	Patient 51	−	Patient 41	−	Patient 121	−	Patient 81	Patient 61	−	−	Patient 131	Patient 19
CTC1 mutation	Val665Gly Pro994fs∗	Val665Gly Pro994fs∗	Val665Gly Pro994fs∗	Val665Gly Pro994fs∗	Val665Gly Leu1142His	Val665Gly Arg1195∗	Val665Gly Leu1007fs∗	Ala227Val Pro994fs∗	Ala227Val Pro994fs∗	Ala227Val Ser353fs∗	Ala227Val Ser353fs∗	Lys242fs∗ Arg975Gly	Cys985 del p.Leu1196_Arg1202 del	−	−
Gender	F	F	F	F	F	F	M	M	M	F	M	M	M	M	F
Age at onset	0.5 years	0.9 years	1.5 years	0.5 years	0.6 years	0.5 years	at birth	14 years	0.1 years	1.0 years	5.0 years	at birth	0.9 years	14 years	27 years
Brain calcifications	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+
Leukoencephalopathy	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+
Intracranial cysts	+	+	+	−	−	+	+	−	−	−	−	+	+	+	+
Retinal abnormality	T, AG, C, R	T, C, R	T, R	T, Rb	T, C, R	T, AG, C, R	+b	−	+b	T, C, Rb	Tb	+b	+b	−	−
IUGR	+	+	+	+	+	+	+	+	+	+	+	+	+	NA	−
Osteopenia or fractures	O	O	O, F	O, F	NA	O	NA	O, F	NA	O	O	O, F	−	NA	−
GI bleeding	+	−	+	+	−	−	+	−	−	+	+	+	−	−	−
Liver involvement	−	−	+	+	−	−	−	−	−	−	+	+	−	−	−
Hematologic findings	−	−	AN, TP, mild LP	AN, TP, mild LP	−	−	−	−	−	AN, mild TP	AN, TP, LP	AN, TP, mild LP	−	−	−
Sparse hair	+	+	−	−	+	−	+	−	−	+	+	+	−	−	−
Mucocutaneous features	−	−	−	−	−	−	−	−	−	−	−	−	−	−	+c
Cardiac abnormality	−	ASD	CAF	−	MVSD	−	−	−	−	−	−	−	−	−	−
Cleft lip	−	−	+	−	−	−	−	−	−	+	−	−	−	−	−
Neurologic symptoms	S, A, D, SZ, HC, CD	S, D, CD	mild CD	S, A, D, CD	−	S, A, SZ, CD	S, HC, DD	SZ, mild CD	DD, HC	S, D, SZ, CD	S, SZ, CD	S, SZ	S, A	S, A, SZ, CN,CD	S, CN, mild CD
Current status (age)	dead (6 years old)	alive (11 years old)	dead (18 years old)	dead (13 years old)	alive (4 years old)	alive (15 years old)	dead (1.9 years old)	alive (22 years old)	dead (1.5 years old)	dead (22 years old)	dead (16 years old)	dead (17 years old)	alive (3.8 years old)	dead (43 years old)	alive (33 years old)

The following abbreviations are used: F, female; M, male; T, telangiectasia; AG, angiomas; C, Coats reaction; R, retinal detachment; IUGR, intrauterine growth retardation; O, osteopenia; F, fractures; NA, data not available; GI, gastrointestinal; AN, anemia; TP, thrombopenia; LP, leukopenia; ASD, arterial septal defect; CAF, coronary artery fistulas; MVSD, multiple ventricular septal defects; S, spasticity; A, ataxia; D, dystonia; SZ, seizures; HC, acute hydrocephalus; CD, cognitive decline; DD, developmental delay; and CN, cranial nerve involvement.

^aThe genotypes are inferred from parental genotypes.

^bAll clinical information for retinal abnormality was not available; additional findings are possible.

^cTwo café au lait spots.

Whole-exome sequencing was performed in two pairs of affected individuals with the most similar phenotypes (F1:II-1 and F6:II-3 had retinal telangiectasias, retinal angiomas, and cerebral cysts; F2:II-1 and F5:II-1 had retinal telangiectasias in the absence of retinal angiomas). Exonic sequences were enriched from three micrograms of genomic DNA with NimbleGen SeqCap EZ Human Exome Library v2.0 (Roche, NimbleGen, USA). Captured DNA was amplified with 12 cycles of PCR. We then used Illumina's Genome Analyzer IIx platform with a 6 bp index read to perform massive parallel sequencing of two paired-end 100 bp reads according to the manufacturer's protocols. We aligned the two 100 bp paired-end sequence reads to the hg19 reference genome by using the Burrows-Wheeler Alignment tool.¹⁰ The Pileup utility from the SAMTOOLS package¹¹ was used for variant calling. Exome sequencing and variant calling were performed at the Institute for Molecular Medicine Finland (FIMM) Technology Centre at the University of Helsinki (for a detailed description of the procedure, see Sulonen et al., 2011¹²). Variant calling resulted in more than 80,000 single-nucleotide variants (SNVs) and 12,000 small insertion-deletion (indel) variants in each individual (Table S1, available online). Of the called SNVs, fewer than 6,000 in each individual were not present in the dbSNP database (build 135). For whole-exome data analysis, we assumed autosomal-recessive inheritance and the presence of one or more deleterious variants in the same gene in all four affected individuals. We considered missense, nonsense, frameshift, and essential splice-site variants as deleterious. We found that no genes with homozygous mutations fulfilled our criteria. However, three genes with compound heterozygous mutations and one gene with a combination of homozygous and compound heterozygous mutations in all four individuals were identified (Table S2).

CTC1 (MIM 613129), which encodes the CTS telomere maintenance complex component 1, was considered as the primary candidate on the basis of its putative function and the absence of the identified variants in exomes of Finnish individuals sequenced at the FIMM Technology Centre (H. Almusa, personal communication). The main transcript of CTC1 is composed of 23 exons, and it encodes a 1,217 aa polypeptide (Figure 2; RefSeq accession numbers NM_025099.5 and NP_079375.3). All four individuals (Figure 3) were heterozygous for a c.1994T>G (p.Val665Gly) missense substitution, two (F1:II-1 and F2:II-1) were heterozygous for a c.2831delC (p.Pro944Leufs^∗7) mutation causing a frameshift and a premature stop codon, one (F5:II-1) was heterozygous for a 2 bp indel (c.3425_3426delTCinsAT) causing a missense substitution (p.Leu1142His), and one (F6:II-3) was heterozygous for a c.3583C>T (p.Arg1195^∗) nonsense mutation (Figures 2 and 3). Conventional Sanger sequencing confirmed the identified mutations (Figure S1). Genotyping of available family members (Figure 3) confirmed that their segregation was consistent with autosomal-recessive inheritance.

Figure 2.

Schematic Structure of CTC1 and the Encoded Protein with Relative Positions of the Mutations

The exons in CTC1 are shown to scale as numbered boxes. The untranslated regions are depicted in gray. The introns are shown as lines and are not to scale. In CTC1, the predicted OB folds are shown as gray boxes. The N-terminal 700 aa region was previously shown to be involved in DNA binding, which is possibly mediated by the two OB domains, and the following C-terminal region in STN1 binding.¹⁴

Figure 3.

Twelve Pedigrees with CRMCC-Associated CTC1 Mutations

Symbols filled with black indicate individuals for whom clinical data were available for the diagnosis of CRMCC. The affection status of individuals F6:II-1 and F9:II-1 (symbols filled with gray) is unknown; their intrauterine growth was severely retarded, they were born prematurely at gestational weeks 36 and 31+3, respectively, and they died soon after birth but were not investigated further. Individual F6:II-2 had clinically definite CRMCC, but his detailed clinical data and DNA were not available for this study. The identified mutations at the protein level are shown below each genotyped individual. DNA samples were available for all individuals with CTC1 alleles indicated, except for F4:II-2 and F7:II-3, from whom only clinical information was available and for whom the genotypes (in parentheses) were inferred from parental genotypes. “wt” denotes no mutation.

Next, we used classical Sanger sequencing to sequence the coding region of CTC1 in the remaining DNA samples of our entire CRMCC cohort containing nine affected individuals (in eight families). In addition, only parent DNA was available for two CRMCC-affected individuals, F4:II-2 and F7:II-3, and we sequenced it to infer the genotypes of these individuals (Figure 3). The CTC1 mutations identified were also sequenced in healthy family members. The primer sequences are given in Table S3, and the PCR conditions are available from the authors by request. Two compound heterozygous mutations were identified or inferred in nine affected individuals in eight families (Figure 3 and Figure S1), but no mutations were identified in F13:II-3 and F14:II-3. Five individuals carried either just one or both of the c.1994T>G (p.Val665Gly) and c.2831delC (p.Pro944Leufs^∗7) mutations identified in whole-exome sequencing. In addition, seven more mutations were identified: two nucleotide substitutions leading to a missense alteration, three deletions creating a frameshift, and two in-frame deletions leading to the removal of a cysteine residue and to a 7 aa deletion (Figure 3 and Figure S1). The 11 identified mutations occurred in 8 of 23 exons in CTC1 (Table 2 and Figure 2).

Table 2.

Summary of the Identified CTC1 Mutations

Nucleotide Change	Exon	Amino Acid Change	PolyPhen Estimate
c.680C>T	5	p.Ala227Val	possibly damaging
c.724_727delAAAG	5	p.Lys242Leufs∗41
c.1058delC	6	p.Ser353Leufs∗14
c.1994T>Ga	12	p.Val665Gly	probably damaging
c.2831delCa	17	p.Pro944Leufs∗7
c.2923A>G	17	p.Arg975Gly	probably damaging
c.2954_2956delGTT	18	p.Cys985del
c.3019delC	19	p.Leu1007Cysfs∗62
c.3425_3426delTCinsATa	22	p.Leu1142His	probably damaging
c.3583C>Ta	23	p.Arg1195∗
c.3586_3606delb	23	p.Leu1196_Arg1202delc

^aMutations found in whole-exome sequencing.

^bThe deleted nucleotide sequence is 5'-TTGTCCTGCCTTTCTATCCGA-3'.

^cThe deleted amino acid sequence is Leu-Ser-Cys-Leu-Ser-Ile-Arg.

The 11 mutations were absent from both the Exome Variant Server of the National Heart, Lung, and Blood Institute (NHLBI) Exome Sequencing Project (ESP, Seattle, WA) database, which includes data of over 5,000 exomes, and the exomes of 367 Finnish individuals sequenced at the Wellcome Trust Sanger Institute (K. Rehnström and A. Palotie, personal communication). By Sanger sequencing 90 Finnish population controls, we found one heterozygous carrier of c.2831delC (p.Pro944Leufs^∗7) and one of c.3019delC (p.Leu1007Cysfs^∗62). Therefore, some mutations might occur, most likely as a result of a founder effect, at a low frequency in the Finnish population, whereas some might be private mutations. Even if the observed frequencies in controls were much lower than those estimated for the diseases of the so-called Finnish disease heritage,¹³ it is somewhat surprising that no individuals homozygous for the two most frequently encountered alterations (p.Val665Gly and p.Pro944fs^∗) were identified. Whether such individuals exist and present with a typical CRMCC or a different phenotype remains to be explored by molecular analysis of further patients.

CTC1 is a component of the trimeric CST complex composed of CTC1, OBFC1 (STN1, oligonucleotide/oligosaccharide-binding fold containing 1 [MIM 613128]), and TEN1 (telomerase capping complex subunit homolog [S. cerevisiae (MIM 613130)]).¹⁴ The CST complex binds to single-stranded DNA (ssDNA) in a sequence-independent manner and associates constantly with a fraction of telomeres, thus probably participating in the protection of telomeres from lethal DNA degradation.¹⁴ CTC1 is predicted to contain three oligosaccharide-binding (OB)-fold domains (Figure 2), suggested to be involved in ssDNA binding.¹⁴ The two N-terminal OB-fold domains mainly contribute to the DNA-binding activity of CTC1, whereas the C-terminal 517 amino acids, which harbor the third OB-fold domain, are responsible for the CST complex formation through interaction with STN1 (Figure 2).¹⁴ In F13:II-3 and F14:II-3, who had remained negative for CTC1 mutations, we also sequenced the coding region of the two other CTS-complex-encoding genes, STN1 and TEN1 (primers listed in Table S4). No mutations were identified.

CTC1 is evolutionarily highly conserved and has orthologs in various plant and animal species, notably in vertebrates.¹⁵ Interestingly, no orthologs have been identified in C. elegans or Drosophila. The orthologs in primates are approximately 97% identical to the human protein, whereas the dog, mouse, and zebrafish orthologs are 76%, 69%, and 32% identical, respectively. The cysteine at position 985, deleted in a Portuguese CRMCC-affected individual (F12:II-1 in Figure 3), as well as three of the four amino acids that were affected by missense substitutions and predicted to be possibly or probably damaging by the PolyPhen¹⁶ analysis, are highly conserved in mammals (Figure 4). The Arg975Gly variant shows less conservation (Figure 4) but is predicted to be probably damaging. The p.Ala227Val change affects the first OB-fold domain and might interfere with DNA binding (Figure 2), whereas the mechanism by which the p.Val665Gly mutant results in impaired CTC1 is less obvious because the mutation does not affect a known functional domain. The last two missense mutations, as well as the cysteine deletion and the 7 aa deletion, are located in the C-terminal region critical for STN1 binding (Figure 2) and are thus predicted to affect the CST-complex formation. The frameshift and premature-stop-codon mutations might result in nonfunctional truncated proteins, protein instability, or mRNA degradation through nonsense-mediated decay.¹⁷ If translated, they would probably result in deficient CST-complex formation.

Figure 4.

Conservation of the CTC1 Amino Acid Sequence among Different Species

Amino acid sequence alignments around the single amino acids affected by five CRMCC mutations are shown for selected species. The figure is modified from data displayed in Ensembl Comparative Genomics Gene Tree alignment for CTC1. The amino acids affected by a missense mutation or a 1 aa deletion in the individuals with CRMCC are highlighted with black. Mutation descriptions are indicated under the sequences. PolyPhen¹⁶ analysis predicted the p.Ala227Val mutation to be possibly damaging and predicted the three other missense mutations to be probably damaging.

Compatible with autosomal-recessive inheritance, no clinical findings typical of CRMCC were found in heterozygous CTC1 carriers. This indicates that one normal copy of CTC1 is sufficient enough for the normal function of CTC1 and the CST complex. Interestingly, individuals with CRMCC in 10 of the 12 mutation-positive families (Figure 3) carried a missense substitution on one allele, and the other allele harbored a frameshift or a premature-stop-codon mutation, which are both predicted to lead to a truncated protein or significant reduction or even total lack of the mutant protein. Similar to missense mutations, the two in-frame deletions in individual F12:II-1 are likely to be translated to protein products with altered functional properties. Only one CRMCC-affected person (F5:II-1) carried two missense mutations (Figure 3). Because no individuals had two frameshift or nonsense mutations, it appears likely that such a combination would either be lethal or result in a different phenotype. In light of the results of experiments in which partial knockdown of CTC1 in mammalian cells disturbed the cell cycle and caused genomic instability and activation of a DNA-damage response,¹⁵ lethality would not be unexpected because such phenomena in a developing embryo would hardly allow survival.

Regarding the differences in phenotype, we either identified or inferred compound heterozygous CTC1 mutations in 13 individuals with CRMCC out of the 15 who were analyzed. CTC1 mutations were identified in all individuals who had onset in childhood and shared a full CRMCC phenotype, including retinal vascular abnormalities, cerebral calcifications, and leukoencephalopathy with or without intracranial cysts (Table 1). On the contrary, CTC1 mutations were identified in only one (F8:II-1) of three individuals with late-onset CRMCC; these individuals were all devoid of clinical retinal abnormalities. The two individuals without CTC1 mutations nevertheless had intracerebral calcifications and cysts but no systemic findings. This suggests that CTC1 is not responsible for the adolescent- or adult-onset LCC phenotype. We are not able to conclude, however, that this would also hold true for the childhood-onset LCC phenotype⁴ because individuals with such a phenotype were not encountered in our study. Interestingly, the CTC1 mutation-positive individual (F8:II-1) with late disease onset had a mild CRMCC phenotype, and although he did not have cerebral cysts and retinal changes, he had suffered from intrauterine growth retardation and had skeletal abnormalities. We thus suggest that systemic findings are important in diagnosing CRMCC and in selecting individuals for CTC1-mutation analysis. Individuals with isolated cerebral findings are not as likely to harbor mutations in this gene. The wide spectrum of clinical manifestations observed in CRMCC is not surprising considering the predicted function and wide expression of CTC1 in human tissues as determined from data available in expression databases (NCBI Gene Expression Omnibus [GEO] and BioGBS). Expression of CTC1 in endothelial cells is in line with the hypothesis that microangiopathy might be a primary underlying pathological abnormality in CRMCC.¹

Finally, there are several overlapping features between CRMCC and disorders in which telomere biology is known to be affected. These are most obvious in Revesz syndrome (RS, [MIM 268130]), a severe and rare variant of dyskeratosis congenita (DC [MIM 127550, 224230, 305000, 613987, 613988, 613989, and 613990]). RS is caused by heterozygous mutations in TINF2,^18,19 which encodes TRF1-interacting nuclear factor 2 (MIM 604319), a major component of the telomere protecting shelterin complex.²⁰ In addition to classical DC manifestations (dysplastic nails, oral leukoplakia, abnormal skin pigmentation, and bone-marrow failure), RS is characterized by intrauterine growth retardation, retinal telangiectasias, exudative retinopathy, and intracranial calcifications;^18,19 such symptoms are strikingly similar to those produced by CRMCC. However, retinal angiomas and cerebral cysts are not characteristic of RS, and cerebellar hypoplasia, which is typical for RS, is not a feature of CRMCC. It is interesting, given that RS and CRMCC both result from mutations in a gene responsible for telomere maintenance, that one individual with RS was reported to have gastrointestinal bleeding, portal hypertension, and hepatopulmonary syndrome,²¹ which are prominent features in some of our CTC1-mutation-positive individuals.

Unlike the RS phenotype, DC phenotypes have seldom been shown to include intracranial calcifications.²² However, retinal findings are fairly frequent. In a survey of 28 individuals with DC, 21% were found to have retinal-pigment-epithelium changes, neovascularization, vascular sheathing, exudative retinopathy, or retinal detachment.²³ Unlike RS, the other known subtypes of DC are caused by mutations in genes that code for either telomerase components (telomerase RNA component, TERC [MIM 602322] and telomerase reverse transcriptase, TERT [MIM 187270]), which are responsible for proper telomere replication and maintenance of chromosome ends,²⁴ or telomerase-associated proteins (dyskeratosis congenita 1, dyskerin, DKC1 [MIM 300126], ribonucleoprotein homolog [yeast] NHP2 [MIM 606470], ribonucleoprotein homolog [yeast] NOP10 [MIM 606471], and WD repeat containing, antisense to TP53, WRAP53 [MIM 612661]).^19,25 A mutation in one of the first five genes or TINF2 is detected in 50%–60% of individuals with DC.¹⁹ Mutations in CTC1 might need to be excluded in the remaining individuals with DC because of an overlap in some retinal and associated findings, such as dysplastic nails, abnormal skin pigmentation, and bone-marrow depression, between DC and CRMCC. Because of the phenotypic overlap, we also searched our whole-exome data for additional mutations in DC-associated genes, but none were observed.

DC-affected individuals who have mutations in telomerase and shelterin-complex components have abnormally short telomeres.^19,26 Previous studies indicated that, in general, the telomere length does not seem to directly correlate with the severity of the DC phenotype,^19,26 but an association between the shortest telomeres and the most severe variants of DC was recently observed.²⁷ Impaired telomerase function resulting in telomere shortening also plays a role in age-related bone loss in telomerase-deficient mice,²⁸ which show an osteoporotic phenotype that involves kyphosis and deteriorated bone architecture mainly in the metaphyseal areas. These features resemble the skeletal characteristics of our cases.⁶ To investigate whether telomere length is affected in CRMCC, we used a quantitative PCR (qPCR)-based method,²⁹ as described before,^30,31 with modifications that are described in Figure 5. Using leukocyte DNA, we determined the length of telomeres from seven individuals with CRMCC (F1:II-1, F2:II-1, F3:II-1, F5:II-1, F6:II-3, F8:II-1, and F11:II-1), nine heterozygous carrier parents, and 70 control individuals. In contrast to the findings for individuals with DC, we did not observe any difference in telomere length between individuals with CRMCC and age-matched controls. We also did not observe a difference between heterozygous carriers and age-matched controls (Figure 5). However, as expected, telomere length in the younger age groups was longer (Figure 5). These data imply that telomere length is not drastically affected in CRMCC, although this should be studied in larger datasets or with other methods.

Figure 5.

Telomere Length in CRMCC-Affected, Heterozygous Carrier, and Control Individuals

Relative telomere length was determined for seven CRMCC-affected individuals (ages 1–16 years) and their 41 controls (ages 7–21years; age matched +/− 5 years) as well as for nine heterozygous-mutation carriers (unaffected parents, ages 30–47 years) and their 29 controls (ages 30–52 years; age matched +/− 5 years). For telomere-length measurement, leukocyte DNA of selected individuals, matched for the DNA-extraction method, was studied by a quantitative PCR-based method,²⁹ as described before,^30,31 with small modifications. PCR reactions were run with a Biorad CFX384 PCR machine controlled by the CFX Manager 2.0 software. All samples were analyzed in triplicate. Quality control was performed as described previously.³¹ Data analysis was carried out in Microsoft Office Excel 2007. Samples were removed from the data analysis if their values deviated more than 3 standard deviations from the mean (N = 2). No significant differences were observed between affected or carrier and control individuals (Student's t test). The error bars represent standard deviation.

It is also possible that CTC1 mutations do not directly affect telomere length, but rather affect the biology of telomeres through some other mechanism. This hypothesis is supported by CTC1-null mutant experiments in Arabidopsis and knockdown experiments of CTC1 in human cell lines.^15,32 Both experiments showed dysregulated telomeres with genomic instability. In Arabidopsis-null mutants, a severe telomere deprotection phenotype accompanied by a rapid onset of developmental defects and sterility was observed.¹⁵ The knockdown of CTC1 expression in human cell lines has indicated both sporadic telomere loss and telomeres with multiple hybridization signals.^15,32 To further study telomere integrity in individuals with CRMCC, we applied a fluorescence in situ hybridization (FISH)-based method with the use of a Cy3-labeled telomere (5'-TTAGGG-3')₃ probe to evaluate the presence of telomere aberrations in metaphase spreads of cultured lymphocytes of two individuals with CRMCC (F5:II-1 and F6:II-3) and two controls (Figure S2). Contrary to CTC1-knockdown findings,^15,32 we did not observe a significant difference between CRMCC-affected and healthy individuals regarding the presence of signal-free ends or the presence of telomeres with multiple hybridization signals (Figure S2). Although limited to the investigation of only two CRMCC-affected individuals, these findings, in line with our qPCR experiments, do not imply severely compromised telomere integrity and might further insinuate more complex underlying biology.

Because the colocalization of the CST complex and telomeres is only a partial one and because CST binds to ssDNA in a sequence-independent manner, it has been suggested that the CST complex has a more general role in DNA metabolism, rather than one limited to telomeric sites.¹⁴ Experiments in mammalian cells have implied that CTC1 plays a role in the stimulation of DNA polymerase α-primase activity by STN1.³³ Recently, studies on Xenopus laevis oocytes showed that CST is involved in the priming of DNA synthesis on the ssDNA template.³⁴ Thus, the role of these processes in the pathogenesis of CRMCC also needs to be explored through further experiments.

In conclusion, we here provide data showing that compound heterozygous CTC1 mutations underlie childhood-onset CRMCC with cerebral and retinal vascular abnormalities and might account for some cases of late-onset disease with a normal fundus appearance. Detection of CTC1 mutations will aid the diagnosis and classification of CRMCC and any related molecularly undefined disorders, including those classified as DC variants, which pose challenges because of the wide spectrum of clinical manifestations. Our data provide the initial step toward understanding the molecular pathogenesis of CRMCC and investigating whether disturbed telomere maintenance or DNA metabolism more generally accounts for the multitude of symptoms—some of which are developmental—associated with the defective CTC1 function.

Web Resources

The URLs for data presented herein are as follows:

• BioGBS, http://biogps.org

• dbSNP135, http://www.ncbi.nlm.nih.gov/snp/

• Ensembl Genome Browser, http://www.ensembl.org/index.html

• Exome Variant Server, NHLBI Exome Sequencing Project (ESP), Seattle, WA,

• http://evs.gs.washington.edu/EVS/

• Interactive Genomics viewer, http://www.broadinstitute.org/igv/

• NCBI GEO Profiles, http://www.ncbi.nlm.nih.gov/geoprofiles

• Online Mendelian Inheritance in Man (OMIM), http://www.omim.org

• PolyPhen, http://genetics.bwh.harvard.edu/pph/

Note Added in Proof

While we were revising our manuscript, a paper describing CTC1 mutations in individuals with Coats plus appeared: Anderson, B.H., Kasher, P.R., Mayer, J., Szynkiewicz, M., Jenkinson, E.M., Bhaskar, S.S., Urquhart, J.E., Daly, S.B., Dickerson, J.E., O'Sullivan, J., Leibundgut, E.O. et al. (2012). Mutations in CTC1, encoding conserved telomere maintenance component 1, cause Coats plus. Nat Genet. Published online January 22, 2012. doi: 10.1038/ng.1084.