Investigation of Chromosome X Inactivation and Clinical Phenotypes in Female Carriers of DKC1 Mutations

Abstract

Dyskeratosis congenita (DC) is an inherited bone marrow failure and cancer susceptibility syndrome caused by germline mutations in telomere biology genes. Germline mutations in DKC1, which encodes the protein dyskerin, cause X-linked recessive DC. Due to skewed X-chromosome inactivation, female DKC1 mutation carriers do not typically develop clinical features of DC. This study evaluated female DKC1 mutation carriers with DC-associated phenotypes to elucidate the molecular features of their mutations, in comparison with unaffected carriers and mutation-negative female controls. All female DKC1 mutation carriers had normal leukocyte subset telomere lengths and similarly skewed X-inactivation in multiple tissue types, regardless of phenotype. We observed dyskerin expression, telomerase RNA accumulation, and pseudouridylation present in all mutation carriers at levels comparable to healthy wild-type controls.

Our study suggests that mechanisms in addition to X chromosome inactivation, such as germline mosaicism or epigenetics, may contribute to DC-like phenotypes present in female DKC1 mutation carriers. Future studies are warranted to understand the molecular mechanisms associated with the phenotypic variability in female DKC1 mutation carriers, and to identify those at risk of disease.

INTRODUCTION

Dyskeratosis congenita (DC) is an inherited bone marrow failure and cancer susceptibility syndrome caused by germline mutations in telomere biology genes. It may present with the mucocutaneous triad of nail dystrophy, reticular skin pigmentation, and oral leukoplakia, but wide clinical variability exists [1, 2]. Patients with DC are at risk of bone marrow failure, vascular complications, pulmonary fibrosis, and other medical complications. They are also at high risk of developing myelodysplastic syndrome, leukemias, and squamous cell carcinomas of the head and neck [3]. Leukocyte telomere lengths less than the first percentile in age-matched populations, as measured by flow cytometry with fluorescence in situ hybridization (flow FISH), are consistent with a diagnosis of DC in the presence of other clinical manifestations [4, 5]. Germline mutations in the X-linked gene DKC1, which encodes the protein dyskerin, were the first to link telomere biology with DC (Knight, et al 1999). Since then, ten other genes integral to telomere function and maintenance have been uncovered to be causative of DC with autosomal dominant (TERC, TERT, TINF2, RTEL1, and ACD) and autosomal recessive (NOP10, NHP2, WRAP53, RTEL1, TERT, ACD, CTC1, and PARN) inheritance patterns; de novo mutations have also been reported [1, 2, 6, 7].

Dyskerin is a highly conserved and essential protein that binds to the H/ACA box in human telomerase RNA (TER) and maintains in vivo stability of TER [8]. After assembly with telomerase reverse transcriptase (TERT), the TERT-TER-dyskerin complex travels to chromosome ends and adds telomeric TTAGGG repeats during the replication phase of the cell cycle. DKC1 mutations result in a dysfunctional dyskerin protein [8, 9].In comparison with wild-type (WT) cells, human cells with dyskerin mutations have a two to five-fold reduction in TER levels, resulting in deficient telomerase activation and defective telomere maintenance [10]. Co-expression of recombinant TERT and TER can restore telomerase activity and telomere maintenance in DKC1 mutant patient fibroblasts without correcting for dyskerin sequence and function [10, 11].

X chromosome inactivation (XCI) is the gene dosage compensation process by which one of the two copies of the X chromosome in females is inactivated. Heterozygous female carriers of DKC1 mutant alleles are typically protected from X-linked DC (X-DC) clinical manifestations through skewed XCI. After gastrulation, cells expressing the DKC1 WT allele have a survival advantage over their counterparts expressing the mutant allele [12, 13]. This growth selection likely results in the depletion of cells expressing the mutant dyskerin allele and thus may explain the lack of DC phenotypes in female carriers. However, DC-like premature aging symptoms, such as early gray hair and delayed wound healing have been reported in a small subset of female DKC1 mutation carriers, the biological mechanism of which is not clearly understood [13, 14].

Recent findings suggest that variable XCI skewing patterns, such as germline mosaicism, could be the cause of clinical manifestations and disease variability in carrier females of X-linked neurological diseases [15]. Similarly, an alternative explanation for the observed DC-like symptoms in DKC1 mutation carriers is that this mosaic XCI skewing pattern could be specific to tissues or organ compartments where clinical features were observed [12, 14, 16]. Additionally, XCI status in blood cells may not be representative of XCI status in tissues where cell lineage representation is usually invariable, such as in the epithelial compartments. However, neither of these hypotheses has yet been fully supported by direct evidence.

In this study, we conducted systematic clinical and biological evaluations of two female DKC1 mutation carriers with phenotypes suggestive of DC, and compared them with an extended panel of clinically unaffected female DKC1 mutation carriers and healthy mutation-negative individuals. We measured XCI status in different cell types, TER stability, total pseudouridine levels, and telomere length in order to better understand the molecular etiology of DC-like clinical manifestations in female DKC1 mutation carriers.

MATERIALS AND METHODS

Study participants

Participants in this study are enrolled in the Institutional Review Board-approved longitudinal cohort study at the National Cancer Institute (NCI) entitled “Etiologic Investigation of Cancer Susceptibility in Inherited Bone Marrow Failure Syndromes” (www.marrowfailure.cancer.gov, NCI 02-C-0052, ClinicalTrials.gov Identifier: NCT00027274) [17]. Affected individuals and their family members completed comprehensive family history and medical history questionnaires. We conducted detailed medical record reviews, collected biospecimens, and performed thorough clinical evaluations of affected individuals and their relatives at the NIH Clinical Center [17]. Whole exome sequencing was performed as previously described [18]. Clinical genetic testing was performed to confirm research-based genetic testing results. The dyskerin mutations involved in this study are illustrated in the schematic of dykserin functional domains in Fig S1.

Cell models

EBV-transformed lymphoblastoid cell lines were derived from female DKC1 mutation carriers and DKC1 mutation negative (wild-type, WT) control subjects. Another EBV-transformed lymphoblastoid cell line (GM03650, heterozygous for DKC1 T66A mutation), was obtained from Corielle Cell Repository. GM03650 was immortalized by retroviral vector-mediated expression of TERT. All cell lines were cultured in RPMI 1640 medium (Gibco/BRL) with 10% fetal bovine serum (FBS) and were maintained at 37 ˚C with 5% CO2.

DNA and RNA isolation

Genomic DNA from whole blood, buccal cells, and fibroblasts was extracted by QIAamp DNA Mini Kit (Qiagen, Valencia, CA). Total RNA from EBV-transformed cells was extracted with Trizol (Invitrogen, Carlsbad, CA). All purified nucleic acids were quantified by Nanodrop Spectrophotometer (Nanodrop Technologies, Oxfordshire, UK).

Telomere length measurement

Telomere length was measured by two different methods: 1) flow FISH in 6-panel leukocyte subsets using previously published protocols [4, 19] and 2) Southern blot-based terminal restriction fragment analysis (TRF) to determine average telomere length [11]. Data from TRF were analyzed by ImageQuant Software (GE Healthcare, Buckinghamshire, UK).

XCI status measurement by the human androgen receptor (HUMARA) assay

XCI status at the DNA level was measured with slight modification to the standard protocol [20]. The highly polymorphic CAG trinucleotide repeat in the promoter area of the androgen receptor was used to study the XCI pattern. In brief, genomic DNA was digested with methylation-sensitive restriction endonuclease HpaII (New England Biolabs, Beverly, MA). The CAG repeat region from digested and undigested DNA samples were PCR amplified with the previously described primers [20] and the PCR products were resolved by denaturing PAGE electrophoresis. The forward primer was end-labeled with radioisotope. The products were visualized by exposure to a storage phosphor screen and scanned with a Typhoon Imager (GE Healthcare, Buckinghamshire, UK).

Direct sequencing of expressed dyskerin allele by RT-PCR

Allelic expression of the DKC1 gene was studied using reverse transcription-PCR (RT-PCR) consisting of 1 μl of total RNA, 1 μl of SuperScript III Reverse Transcriptase (Life Technologies, Gaithersburg, MD), 4 μl of 5x first-strand buffer, 2 μl of 100mM DTT, 2 μl of 10mM dNTP mix, and 1 μl of 25uM specific reverse primer in a total reaction volume of 20 μl. The reaction was incubated at 42 ˚C for an hour before the inactivation of the enzyme at 70 ˚C for 15 minutes. Half of the first-strand cDNA was used for PCR amplification with specific primers. Genomic DNA was PCR amplified in parallel. The PCR products from both cDNA and DNA were resolved by agarose gel electrophoresis, purified by QIAquick Gel Extraction Kit (Qiagen, Valencia, CA), and sequenced (Genewiz). Sequences for the primers are listed in Tables SI and SII.

Protein expression measurement by western blot

30 μg whole cell extracts were resolved in non-continuous Tris-glycine sodium dodecyl sulphate gels. The resolved protein samples were transferred to polyvinylidene fluoride (PVDF) membranes (GE HealthCare Life Sciences, Piscataway, NJ). The blots were incubated with anti dyskerin polyclonal antibodies (1:1,000, Santa Cruz Biotech) and anti-β-actin monoclonal antibody (1:40,000, Sigma). Protein signals were labeled with Alexa Fluor 680 Dye (Thermo Fisher Scientific, Rockford, IL), detected by Licor Odyssey CLx Infrared Imaging System and quantified by ImageJ software.

TER copy number measurement by competitive RT-PCR

TER was quantified by a previously published protocol [11, 21]: 0.5 μg of total RNA from each sample was mixed with sample-specific ranges of TER competitors in a 2-fold dilution series. Following RT-PCR and product resolution by non-denaturing PAGE electrophoresis, the products could be visualized as the forward primer in the PCR was labeled with a CY5 fluorophor tag. The fluorescent image was scanned by a Typhoon imager, and the densitometry readings were quantified with ImageQuant software.

Pseudouridine level measurement by HPLC

The analytical method for the determination of pseudouridine in RNA was modified based on previous reports [22, 23]. Quantitative analysis was performed by HPLC-UV system consisting of a Waters 2695 HPLC system and a Waters 2996 UV detector (Waters, Milford, MA). Data acquisition and processing was performed by Empower software. The chromatography was performed on a Waters Nova-Pak C18 column 3.9 mm x 300 mm, 4 μm.

Statistics

All data were analyzed by GraphPad Prism software (GraphPad Software, San Diego, CA). Error bars denote SEM. The student’s t-test was used for two group comparisons, and the one-way ANOVA with post-hoc Bonferroni correction was used to adjust for multiple comparisons. Differences were considered significant at p < 0.05.

RESULTS

Study participants

We evaluated two female DKC1 mutation carriers who have phenotypic features overlapping with DC (Table I).

Table I.

Characteristics of female heterozygous DKC1 mutation carriers and healthy controls.

Clinical status	Participant ID	Age at sample collection (years)	DKC1 mutation
Affected carrier	NCI 219–2	15	ΔE35
	NCI 288–4	65	L54V
Unaffected carrier	NCI 96–2	55	K390Q
	NCI 103–2	59	ΔL37
	NCI 103–3	30	ΔL37
	NCI 106–4	36	K314R
	NCI 167–4	51	R322Q
	NCI 219–5	40	ΔE35
	GM03650	47	T66A
Healthy wild-type control	NCI 140	49	None
	NCI 156	31	None
	NCI 160	31	None
	NCI 204	50	None
	NCI 228	33	None

Subject NCI-288–4, a female carrier of the DKC1 L54V mutation, has alopecia and ridged nails. She comes from family NCI-288, which consists of a male proband with classic features of DC (Fig 1A). Genetic testing for known DC genes revealed a single mutation in DKC1 (p.L54V). His mother (NCI-288–3), who was confirmed to be a carrier of the same DKC1 mutation (p.L54V), had multiple phenotypic features consistent with DC including aplastic anemia, non-alcoholic liver cirrhosis, abnormal nails, and early tooth loss. She died secondary to end stage liver disease and hepatic encephalopathy at 65 years of age. Her sister (NCI-288–4) was evaluated as part of this study.

Figure 1. Family pedigrees of clinically affected female DKC1 mutation carriers.

A. Family NCI 288 with affected proband NCI 288–1, his affected carrier mother NCI 299–3, and affected carrier maternal aunt NCI 288–4

B. Family NCI 219 with affected proband NCI 219–2 and her unaffected carrier mother NCI 219–5C.

C. Clinical manifestations of DC-associated phenotypes in female DKC1 mutation carrier NCI 219–2 i and ii. Characteristic skin pigmentation iii and iv. Characteristic nail dysplasia

Proband NCI-219–2 is a teenage girl whose brother and maternal uncle were both affected with DC. Her brother (NCI-219–1) died of post- hematopoietic stem cell transplant complications at age 3 years. She has two unaffected sisters (Fig 1B). The proband has nail dystrophy since birth, and skin hyperpigmentation and early graying of hair since the age of 10 years, all of which are phenotypes associated with DC (Fig 1C). Whole exome sequencing (WES) performed on the proband and her unaffected mother revealed a heterozygous mutation causing a single amino acid deletion in the coding sequence of DKC1 (p.ΔE35) that was clinically validated to be present in both femlaes. Genetic testing of the unaffected sisters revealed that they do not harbor the mutation. WES did not reveal mutations in any of the other known DC genes.

In addition to these two female DKC1 mutation carriers with phenotypic features, we also identified six clinically unaffected female DKC1 carriers, including the mother of NCI-219–2, all of whom were immediate family of males with X-linked DC. They served as a comparison group in the clinical and functional evaluations in this study. We also evaluated five healthy age-matched female WT controls who were mutation negative for DKC1 and all other DC-associated genes (Table I).

Leukocyte telomere length

To explore the hypothesis of whether clinical manifestations of DC-associated symptoms in female DKC1 mutation carriers were caused by short telomeres, we measured 6-panel leukocyte subset telomere lengths (granulocytes, total lymphocytes and lymphocyte subsets) of all affected and unaffected female DKC1 mutation carriers and compared them with healthy female controls. Female DKC1 mutation carriers had normal telomere lengths in granulocytes and all lymphocyte subsets, regardless of clinical phenotype, and were comparable to age-matched healthy controls (Fig 2). In families where data were available, flow FISH telomere lengths of the male probands with DKC1 mutations were very short (less than first percentile for age) in all leukocyte and lymphocyte subsets (Fig 2). The average leukocyte telomere length measured by TRF analysis of Southern blots correlated with and confirmed the flow FISH findings in the female DKC1 carriers and healthy female controls (Fig S2).

Figure 2. Mean lymphocyte telomere length in study participants.

Total lymphocyte telomere length measured by flow-FISH is shown. The vertical axis represents telomere length in kilobases. Lines in the figures indicate the first, tenth, 50th, 90th, and 99th percentiles of results from 400 normal control subjects. Data include the telomere length measurements from two DC male probands for comparison.

Evaluation of X Chromosome Inactivation

The modified HUMARA assay for XCI was performed on samples from seven subjects in five X-DC families: families NCI-219 (with DKC1 mutation p.ΔE35), NCI-103 (p.ΔL37), NCI-288 (p.L54V), NCI-96 (p.K390Q) and the p.T66A sample (Table I, Fig 3A). XCI status in DNA samples from both blood and buccal mucosa were tested simultaneously. For the female proband NCI-219–2, the HUMARA assay was also performed on DNA from fibroblasts obtained from skin areas with and without pigmentation changes (Fig 3A). We observed skewed XCI in all samples, regardless of their DKC1 mutations, clinical phenotype, or tissue of origin. In contrast, using the same assay, we observed mosaic XCI patterns in all leukocyte DNA samples from healthy controls (Fig S3).

Figure 3. Skewed X-chromosome inactivation and normal dyskerin protein expression in female DKC1 mutation carriers, regardless of tissue origins or DKC1 mutations.

(A) Skewed X-chromosome inactivation (XCI) patterns were detected in leukocytes and buccal cells for female mutation carriers in DKC1 ΔE35, ΔL37, L54V, T66A, and K390Q families by HUMARA assay. In addition to PBMC, differently pigmented skin samples were collected from the female proband in family NCI-219, DKC1 ΔE35. We tested the XCI status in her PBMC and fibroblasts derived from her pigmented skin (denoted with an asterisk) as well as from normal skin areas. All samples from this individual showed complete skewed X-inactivation. (B) Only wild-type (WT) dyskerin was observed in EBV-transformed lymphoblastoid cells derived from female carriers. The nucleotide sequences of both genomic DNA and complimentary DNA of female carriers are illustrated on top of the chromatogram. DKC1 genotypes at the variant nucleotide(s) for each family are illustrated in parentheses on top of the DNA sequences. (C) A representative western blot for dyskerin; β-actin was used to control the loading. (D) The dyskerin expression in EBV-transformed lymphoblastoid cells was quantified and normalized to their β-actin expression. The bold line and whiskers represent mean dyskerin expression level (normalized to β-actin) ± 1 standard deviation. NS, not significant.

Talebizadeh et al. reported that factors other than the XCI process could affect the in vivo allelic expression of X-linked genes and that evaluation of XCI status at the gene expression level was essential to understand the effects of XCI for each genetic loci [24]. To confirm that the mutant dyskerin allele did not escape XCI, we evaluated dyskerin mRNA expression by sequencing the dyskerin locus from genomic DNA to confirm their DKC1 mutant carrier status, and compared the genomic sequencing data with RT-PCR-sequencing analysis of dyskerin mRNA expression (Fig 3B). Only wild-type (WT) dyskerin mRNA expression was detected in the EBV-transformed lymphoblastoid cells from one affected and four unaffected female heterozygous DKC1 mutation carriers evaluated.

Dyskerin function and protein expression

Loss of stable TER accumulation is a molecular marker of X-linked DC. We evaluated and compared the level of stable TER in female mutation carriers with that of healthy controls in order to assess the ability of dyskerin to maintain TER accumulation (Fig S4A). Our data show that TER levels were significantly higher in both female DKC1 mutation carriers and female DKC1-WT controls than in males with X-linked DC (data from [11]) (one-way ANOVA with Bonferroni post hoc test, p < 0.0001, Fig S4B). However, there were no significant differences in TER expression levels among the healthy WT controls and female mutation carriers (Student’s t-test, p = 0.33) (Fig S4B).

In addition to maintaining TER stability, dyskerin also has an essential function as a pseudouridine synthase in the site-specific modification of ribosomal RNAs (rRNAs) [25, 26], small nuclear RNAs (snRNAs) [27], non-coding RNA (ncRNA) and a subset of mRNA [28]. Previous studies have showed that at least two disease-associated DKC1 mutations are associated with a reduction in total rRNA pseudouridine levels [22]. As several DKC1 mutant alleles in our current study had not been previously evaluated, we quantified pooled pseudouridine levels in total RNA from our panel of female carriers of DKC1 mutations. There were comparable levels between DKC1 mutation carriers and healthy WT controls (Student’s t-test, p = 0.52), suggesting that dyskerin’s pseudouridinylase function is not compromised in female DKC1 mutations carriers (Figs S5A and S5B).

To evaluate whether there were differences in dyskerin protein expression between female DKC1 mutation carriers and their age- and gender-matched healthy WT controls, we measured dyskerin expression levels by western blot analysis, and normalized dyskerin signals to the expression of β-actin, as loading control. No statistically significant difference was observed in normalized dyskerin protein expression level between the two groups (Student’s t-test, p = 0.53) (Fig 3C and 3D).

DISCUSSION

This study revealed that there appears to be normal telomere length maintenance in the examined tissues from all female carriers of DKC1 mutations, including those affected with features of DC and carriers from families with affected males who have very short telomeres. We did not observe mosaic XCI skewing patterns or aberrant dyskerin gene expression levels in female DKC1 mutation carriers. Our data also showed no functional changes in dyskerin protein expression, TER stability and total RNA pseudouridine levels when comparing female DKC1 mutation carriers (affected or unaffected) with healthy WT controls. This is consistent with previous reports of normal telomere biology in female DKC1 mutation carriers [14].

Functional analysis of the mutation p.Leu54Val (in family NCI-288) was previously performed in induced pluripotent stem cells (iPSCs) derived from male probands [29]. In agreement with classical DC, p.L54V-derived fibroblasts and iPSCs were found to harbor reduced TER level, impaired telomerase activity and defective telomere maintenance. The mutation p.ΔE35 in family NCI-219 has never been reported before our study, In this family, two male members displayed classical DC manifestations of immune deficiency and bone marrow failure early in life, and they both died at ages 2 to 3 years. Since no other mutations in coding sequences of known DC genes were found in this family, we infer that p.ΔE35 is responsible for the clinical phenotypes. Notably, both p.ΔE35 and p.L54V are located in a mutation hot spot in exon 3 at the N-terminus of dyskerin coding sequence (Fig. S1). Additionally, a closely related residue (p.ΔL37) had been reported in numerous biochemical studies to encode for classical DC, from our laboratory and others [11, 30].

It remains unclear why certain female DKC1 mutation carriers develop DC-like phenotypes, in the presence of functionally normal dyskerin and normal telomere maintenance. One hypothesis is that secondary skewing of XCI may not apply to parenchymal tissues because their compositions of cell lineages are not as dynamically edited. As a consequence, dysfunctional dyskerin expressing cells may still exist in a subset of parenchymal cells and contribute to the mosaic epithelial symptoms observed [14]. Our evaluation of XCI status in DNA samples collected directly from unprocessed buccal cells (from all DKC1 mutation carriers) and from normal and abnormally pigmented skin from NCI-219–2, an affected carrier, showed completely-skewed XCI (Fig 3A). These data suggested that X chromosome encoding for the mutant DKC1 allele was not expressed in vivo.

We also observed wildtype DKC1 expression (Fig 3B), normal dyskerin expression (Fig 3C and 3D) and function (Fig S4 and S5) with cultured cells derived from primary patient samples. While these data were in complete agreement with the skewed-XCI status and normal telomere length observed with uncultured clinical materials, it is important to note the assays’ limitation: culturing of dermal tissue as fibroblasts, and blood cells as EBV-transformed lymphoblasts may have allowed for growth selection against cells expressing mutant dyskerin; thus, the negative results from our study using fibroblasts and lymphoblasts as sample source could be explained by the selection bias favoring WT dyskerin-expressing cells in ex vivo culture. With the recent developments in in situ PCR with single cell resolution, these targeted collection of clinical samples, such as skin biopsies from affected and unaffected areas, may help to shed light on the potential mosaic dyskerin expression in parenchymal tissues at DKC1 inactivation stages [31] or at allelic expression level [31, 32].

There were no concurrent mutations found in other known DC genes in the affected female carrier that may explain the sporadic appearance of DC-like phenotypes in some but not other DKC1 mutation carriers. However, the lack of overt changes in dyskerin function in affected female carriers could still be explained by the compound presence of genetic and/or environmental modifiers that have not yet been identified to influence telomere biology. Telomere length maintenance is a dynamic process that involves both synthesis and attrition. Weak perturbations of telomere pathway components, such as non-uniform expression of mutant dyskerin in female carriers, could predispose individuals to DC-like symptoms. But other genetic predispositions and/or environmental cues may be needed to induce clinical manifestations [33].

The dyskerin complex has recently been found to potentiate in vitro OCT4/SOX2-mediated transcription in embryonic stem cells and depletion of dyskerin in human fibroblasts resulted in decreased efficiency of iPSC conversion [34]. As pluripotent embryonic stem cells are produced before the gastrulation stage during embryogenesis when XCI is established, female heterozygous DKC1 mutation carriers expressing both DKC1 alleles (WT and mutant) may have differential expression of pluripotency genes in their embryonic stem cells. These changes in early developmental stages may be followed by normal XCI and selection for WT-DKC1 expressing cells. It is possible that mutant dyskerin may affect cellular differentiation and maturation in the female heterozygous mutant carriers.

In summary, we systematically studied dyskerin biology in a panel of female heterozygous carriers of disease-associated DKC1 alleles. We observed extensively skewed XCI patterns in all female carriers, in samples from blood and buccal cells, as well as in patient-derived primary fibroblast cultures and lymphoblastoid cell lines. Exclusive expression of WT-dyskerin in these female subjects may protect them from severe manifestations of DC-associated molecular phenotypes. Female DKC1 mutation carriers had normal telomere lengths, as well as normal dyskerin expression and function. These findings are surprising given the extent of clinical phenotype observed in patients such as NCI-219–2, NCI-288–3 and NCI-288–4. Clinical manifestations of DC phenotypes in female carriers may be a result of multiple factors, including the interactions between mosaic (temporal and positional) expression of mutant DKC1, environmental and behavioural factors that influence the rate of telomere attrition as well as telomere length inheritance from affected family members. Further studies are warranted to elucidate the disease mechanism in female carriers.

Supplementary Material

Table SI

Table SI. The primer pairs for genomic DNA-PCR and RT-PCR.

Table SII

Table SII. Sequences for the primers used in the study

Fig S1

Figure S1. Schematic of dyskerin functional domains. Non-synonymous amino acid (aa) changes in this study are illustrated in red. Studies from orthologs in other species, and from biochemical studies of human dyskerin in cell models, have mapped the following functional features in the dyskerin protein. It has a TruB domain (aa107 – aa247), a PUA domain (aa297 – aa371) and two nuclear localization signal regions (aa11 – aa20, aa446 – aa458). TruB domain (named after an Escherichia coli pseudouridine synthase) serves as the catalytic domain of pseudouridine syntahses. PUA domain (pseudouridine synthase and archaeosine-specific transglycosylase domain) is the putative RNA-binding domain.

Fig S4

Figure S4. Female DKC1 mutation carriers have normal TER levels at steady state. (A) Steady-state TER levels were measured in EBV-transformed lymphoblastoid cells using a previously published competitive RT-PCR assay (see methods for details). Input-tagged competitor RNA competitor RNA copy numbers are indicated for each series. The arrows indicate the approximate competitor inputs where 1:1 ratio of competitor to endogenous TER RT-PCR signals were obtained. For the quantification experiment, each RNA sample was measured first with a 10-fold serial dilution range of input competitor, then repeated with an adequate 2-fold serial dilution range for final copy number calculation. (B)The densitometry quantification data are shown in the dot plot. The bold line and whiskers represent mean TER level (expressed as copy numbers of the competitor RNA) ± 1 standard deviation. NS, not significant. Though TER levels in both DKC1 mutant carriers and WT control samples are highly variable, all of them are higher than that in TERT-expressing fibroblasts from X-DC male patient (data from [11]) (p < 0.0001, one-way ANOVA with post-hoc Bonferroni correction).

Fig S3

Figure S3. Mosaic XCI patterns observed in DNA from leukocytes of healthy DKC1 mutation negative women by the HUMARA assay.

Fig S2

Figure S2. Mean leukocyte telomere lengths in female DKC1 mutation carriers and mutation negative controls. Terminal restriction fragment analysis was used to quantify telomere lengths. The measured signal represents the sum of telomeric and sub-telomeric region. Estimated telomere length was calculated by the weighted average of the lane and is shown at the bottom of the gel. All samples tested here were collected from leukocyte DNA without cell culture. The only exception (buccal cell DNA sample) is labelled with the asterisk. It shows that telomere length in parenchymal tissues is similar to that in blood cells.

A. Female DKC1 mutation carriers

B. Female mutation-negative healthy controls

Fig S5

Figure S5. Female DKC1 mutation carriers have normal pseudouridine modification levels. (A) A representative HPLC of digested total RNA profile: the retention time for each nucleoside is illustrated in the chromatogram, and 7-methylguanosine (7-met G) is used as an internal control. (B) Pseudouridine level in total RNA. Steady-state pseudouridine levels in total RNA from female DKC1 mutation carriers are comparable to those from WT controls. The bold line and whiskers represent mean pseudouridine level in total RNA (normalized to cytosine) ± 1 standard deviation. NS, not significant.