How to cite this article: Morice-Picard F, Cario-André M, Rezvani H, Lacombe D, Sarasin A, Taïeb A. 2009. New clinico-genetic classification of trichothiodystrophy. Am J Med Genet Part A 149A:2020–2030.
Trichothiodystrophy (TTD) is a rare ectodermal disorder first described by Pollitt et al. 1968 and named by Price et al. 1980. The patients usually present with dry and sparse hair. Hair shafts break easily with trauma. The name trichothiodystrophy was proposed to group several phenotypes on the basis of a common deficiency in sulfur proteins of the hair shaft [Price et al., 1980]. Several neuroectodermal manifestations are variably seen in this phenotype including mental retardation, ichthyotic skin, reduced stature, osseous anomalies and hypogonadism but none is a constant trait [Itin and Pittelkow, 1990; Itin et al., 2001].
TTD is inherited as an autosomal recessive trait [Jackson et al., 1974; Price et al., 1980; Howell et al., 1981]. DNA repair defect is present in around 50% of patients, which have been referred to as “photosensitive” even without clear-cut evidence of associated clinical photosensitivity [Nishiwaki et al., 2004]. Mutations in XPD, XPB, p8 have been subsequently found in “photosensitive” TTD patients [Stefanini et al., 1986; Weeda et al., 1997; Giglia-Mari et al., 2004]. Mutations in C7Orf11, encoding TTDN1 of unknown function was found in the group of patients without DNA repair anomalies [Nakabayashi et al., 2005]. However mutations in this gene were excluded in some non-photosensitive patients including those described by Howell et al. 1981 (Sabinas syndrome) and Pollitt et al. 1968 suggesting further genetic heterogeneity in the group without DNA repair anomaly [Nakabayashi et al., 2005]. A genetic classification into three groups can thus be proposed distinguishing a group with DNA repair anomalies (I), a group without DNA repair defect and with TTDN1 mutations (II), and a group without DNA repair defect and without identified genetic basis (III).
XPD, XPB, and p8 are subunits of the transcription/DNA repair factor IIH (TFIIH). TFIIH is a complex consisting of 10 proteins essential for both nucleotide excision-repair (NER) and transcription [Schultz et al., 2000; Coin et al., 2006; Laine and Egly, 2006]. Mutations in subunits associated with TTD destabilize the TFIIH structure and lead to decreased cellular concentrations [Coin et al., 1998; Vermeulen et al., 2000]. The lower amount of TFIIH found in individuals with TTD contributes to a limiting level of transcription of targeted genes and could explain the TTD phenotype including cutaneous and neurological features [Compe et al., 2007].
TTDN1 is a nuclear protein not involved in DNA repair. It has been shown that TTDN1 interacts with polo-like kinase 1 (PLK1), a highly conserved serine-threonine kinase regulating cellular cycle and mitosis [Zhang et al., 2007]. TTDN1 has several phosphorylation sites and is a regulator of mitosis. Interactions between cell cycle regulation and transcription efficiency could explain the TTD phenotype observed in patients with TTDN1 mutations.
In this article, we review patients seen in our Clinical Department and those published to compare the phenotypes in photosensitive and non-photosensitive groups, especially for cutaneous, neurological, osseous and gonadal aspects with the aim to establish genotype–phenotype correlations in TTD.
PATIENTS AND METHODS
We analyzed the clinical condition of TTD patients and their genetic characterization when available, through a literature review. Patients were included if the characteristic hair anomalies were present, including a sulfur deficiency of hair and an abnormal microscopic aspect (trichoschisis and hair-banding under polarized light) allowing a definite diagnosis of TTD. Following a comprehensive literature review, we selected a series of informative features of the TTD phenotype to establish a clinical database, including mental retardation, growth failure, osteosclerosis, gonadal dysfunction, cutaneous changes, and clinical photosensitivity. We looked at the associated genetic status of each patient published in the literature, generally in consecutive reports. For the analysis, two groups where distinguished, namely group A with DNA repair anomalies and group B without DNA repair anomalies and irrespective of the classification in three genetic groups. We compared the frequencies of the selected clinical findings in both groups using the χ2 test (α = 5%).
We reviewed 122 patients with criteria for TTD [Pollitt et al., 1968; Brown et al., 1970; Tay, 1971; Jackson et al., 1974; Arbisser et al., 1976; Jorizzo et al., 1980; Price et al., 1980; Howell et al., 1981; Crovato and Rebora, 1983; Diaz-perez and Vasquez, 1983; Van Neste and Bore, 1983; Happle and Traupe, 1984; King et al., 1984; Lucky et al., 1984; De Prost et al., 1986; Rebora et al., 1986; Stefanini et al., 1986, 1992; Meynadier et al., 1987; Baden and Katz, 1988; Fois et al., 1988; Lehmann et al., 1988; Motley and Finlay, 1989; Van Neste et al., 1989; Broughton et al., 1990; Przedborski et al., 1990; Kousseff, 1991; Savary et al., 1991; Peserico et al., 1992; Rizzo et al., 1992; Sarasin et al., 1992; Alfandari et al., 1993; Calvieri et al., 1993; Hersh et al., 1993; McCuaig et al., 1993; Chen et al., 1994; Feier and Solovan, 1994; Tolmie et al., 1994; Eveno et al., 1995; Lynch et al., 1995; Bracun et al., 1997; Brusasco and Restano, 1997; Malvehy et al., 1997; Schepis et al., 1997; Takayama et al., 1997; Botta et al., 1998, 2002, 2009; Petrin et al., 1998; Foulc et al., 1999; Itin et al., 2001; Toelle et al., 2001; Vermeulen et al., 2001; Viprakasit et al., 2001; Mazereeuw-Hautier et al., 2002; Dollfus et al., 2003; Giglia-Mari et al., 2004; Wakeling et al., 2004; Faghri et al., 2008].
DNA repair analysis data and genetic status was available on 79 patients.
UV-induced DNA repair deficiency was found in 42 (group A). The genetic status was available on 36 patients in group A, including 30 patients with XPD mutations [Crovato and Rebora, 1983; King et al., 1984; Stefanini et al., 1986, 1992; Broughton et al., 1990; Peserico et al., 1992; Chen et al., 1994; Tolmie et al., 1994; Eveno et al., 1995; Takayama et al., 1997; Botta et al., 1998, 2009; Foulc et al., 1999; Vermeulen et al., 2000; Boyle et al., 2008], 4 with p8 mutations [Jorizzo et al., 1980; Giglia-Mari et al., 2004] and 2 with XPB mutations [Sarasin et al., 1992; Weeda et al., 1997]. Six patients were presenting with abnormal DNA repair without molecular characterization.
Thirty-seven patients presented with normal DNA repair (group B). Twenty-eight patients had TTDN1 mutations [Nakabayashi et al., 2005; Botta et al., 2007]. We obtained clinical data on 26 of them [Jackson et al., 1974; Diaz-perez and Vasquez, 1983; Fois et al., 1988; Lehmann et al., 1988; Przedborski et al., 1990; Rizzo et al., 1992; Nakabayashi et al., 2005; Botta et al., 2007]. Normal DNA repair was found in 11 patients, two of them had no TTDN1 mutations [Nakabayashi et al., 2005]. Molecular analysis was not performed in the nine remaining patients.
Nine TTD cases were diagnosed at our institution between 1982 and 2007. Only the seven patients on whom DNA repair analysis data are available are described here.
Patient 1 (TTD1VI)
This boy was the first child born at term to nonconsanguineous healthy parents. Family history was not relevant. Intrauterine growth retardation was noted. Birth weight (BW) was: 2,740 g (−1.5 SD), length (BL) 44 cm (−2 SD). Ichthyosiform lesions of lower legs were noted on the 8th day of life. He was first seen at 8 years for severe psychomotor delay. Neurologic exam showed axial hypotonia with peripherical hypertonia and generalized convulsions. He had large protruding ears and unilateral single palmar crease. Sparse brittle hair was noted. Magnetic resonance imaging (MRI) showed pachygyria and ventricular dilatation. Metabolic investigations including very long chain fatty acid level, lysosomal enzymatic activities, mucopolysaccharides dosage were normal. Hair microscopy showed a tiger-tail banding and biochemical exam displayed a low-sulfur-hair content thus confirming the diagnosis of TTD. UV-induced DNA repair deficiency was found in vitro at about 30–40% of control. Full complementation was observed following transfection with wild-type XPD gene. XPD gene analysis found two deleterious mutations (First allele: p.R722W and second allele p.L461V/716-730del) [Takayama et al., 1997].
Patient 2 (TTD3VI)
This boy was born at term after a normal pregnancy. Family history was uninformative. He was first seen for chronic alopecia. Clinical examination showed ichthyosis most prominent on trunk, dry sparse hair with alopecia. Photosensitivity was noted since age 7 years. A major psychomotor delay was present. Hair analysis showed a tiger-tail pattern under polarized light (Fig. 1). Amino-acid dosage shown diminished sulfur hair content confirming the diagnosis of TTD. Skin histological examination showed a thin granular layer (Fig. 2). Zonular cataract was present. UV-induced DNA repair deficiency was confirmed in vitro. XPD gene analysis found two deleterious mutations (First allele: p.R658H, and second allele p.L461V and Del p.716-730) [Takayama et al., 1997].
Patients 3 and 4
The first child of a first-cousin healthy couple was a boy. He was born at term after an uneventful pregnancy. He was seen at birth, when he presented with congenital ichthyosis (collodion baby) progressing to mild ichthyosis on the trunk (Fig. 3a). Diagnosis of TTD was suspected at age 3 years on the basis of mild ichthyosis of trunk, scalp, palms and soles (Fig. 3b), mild photosensitivity noted after sun exposure and hair macroscopically normal but coarse and with a tiger-tail pattern under polarized light. The diagnosis of TTD was confirmed by the analysis of hair aminoacid content. Minor facial anomalies included broad nasal bridge, apparently low-set abnormaly modelate ears. Growth and psychomotor development were normal. Osseous radiographies were normal. A full blood count was normal. Hemoglobin electrophoresis was normal. IgE were elevated. A UV-induced DNA repair anomaly was detected. Complementation analysis showed for the first time that the DNA repair defect was associated with the XPB gene [Weeda et al., 1997]. Homozygous deleterious mutations in the XPB gene were found. He was seen again at age 22. Photosensitivity had disappeared. Ichthyosis of the flanks was more marked and associated with a palmar hyperkeratosis (Fig. 3c–e). A sensorineural deafness was recently diagnosed.
The second child was a girl. She was born at term with the same presentation of congenital ichthyosis with a favorable outcome (Fig. 4a). The diagnosis of TTD was confirmed by hair microscopy and biochemical analysis. There was no mental or growth delay. She was reevaluated at age 18. Ichthyosis was noted on the trunk with desquamation following sun exposure (Fig. 4b). She had also mild deafness. The two patients have two homozygous mutations in the XPB gene (p.Y119P) [Weeda et al., 1997].
The patient is the first child of a nonconsanguineous healthy couple. He was examined at 15 years for short, sparse hair with alopecia associated with a generalized xerosis and keratosis pilaris. Hair microscopical exam shown trichoschisis associated to a tiger-tail banding under polarized light. Hypogonadism was associated with micropenis and cryptorchidia. He also had leucopenia, microcytosis, myopia, moderate developmental delay but normal growth. Bone X-rays were normal. The diagnosis of TTD was confirmed by hair aminoacid analysis showing a low sulfur hair content. UV-induced DNA repair analyses were normal.
Patient 6 and 7
The boy (No. 6) was the first child of a non-consanguineous healthy couple. Toxaemia was noted during pregnancy. He was born at 37 weeks (birth weight 2,700 g) and was found to have a congenital ichthyosis (collodion baby). He was first seen at age 7 years with his sister who had presented with the same history of toxemia and collodion baby. Both had sparse, brittle, hypopigmented hair. Microscopical and biochemical analysis of hair confirmed a diagnosis of TTD. Hair loss following fever episodes was noted. The girl (No. 7) had early onset insulin-dependent diabetes. Other findings included moderate leuconeutropenia, normal growth and developmental delay. DNA repair studies were normal in both sibs. Both of them had a marked pigmentary dilution (skin and hair) as compared with their parents.
Phenotype Analysis of 79 TTD Patients
The non-cutaneous aspects are summarized in Table II.
Table II. Frequency of the Main Non-Cutaneous Features in TTD Groups A and B
n = 42
n = 37
Failure to thrive
Neurological involvement consists mainly in mental retardation, and more uncommonly in ataxia, spastic paralysis or cerebellar atrophy. Convulsions have rarely been described. MRI may show abnormal white matter aspects. The two groups were similar. For example, mental retardation was observed in 87% of in vitro photosensitive patients (group A) and in 84% of the non-in vitro photosensitive patients (group B).
Growth failure was observed in both groups, usually moderate and rarely severe.
Typical osseous manifestations manifested as axial osteosclerosis with peripherical osteopenia. Osseous manifestations were more frequent in group A, but the difference was not significant (χ2 = 2.62). Osseous anomalies can be asymptomatic and are not often specified in the reports.
Hypoplastic testis, cryptorchidia are observed in males. Hypogonadism has been sometimes substantiated by hormonal dosages (diminished testosterone, elevated FSH, LH levels). A low fecundity rate has been observed in the TTDN1-mutated Amish families suggesting a defect in fertility. This may be the consequence of gonadal dysfunction. Gonadal anomalies were significantly more frequent in group B than in group A (65% in group B vs. 24% in group A; χ2 = 13.69). However the lack of detailed clinical description including gonadal function constitutes a bias in this analysis.
The results are summarized in Table III. Skin anomalies mainly consisting of ichthyosis. An aspect of collodion baby with a favorable course may precede the development of ichthyosis. The descriptions of ichthyosis are often consistent with the vulgaris type with small, white scales of the legs. Other findings include xerosis, palmoplantar keratoderma, atopic dermatitis, follicular keratosis. Pooled data shows a significantly elevated frequency of ichthyosis in group A than in group B (χ2 = 47), and a higher prevalence of neonatal forms (collodion baby) (χ2 = 5.34). Moreover no description of collodion baby was found in patients with TTDN1 mutations [Nakabayashi et al., 2005; Botta et al., 2007]. However, two of our patients were described as mild collodion babies and were subsequently found to have normal DNA repair. Skin changes in group B are most often non-specific, consisting of xerosis, follicular keratosis, and atopic dermatitis.
Table III. Frequency of Cutaneous Anomalies in TTD Groups A and B
n = 42
n = 37
The frequency of ichthyosis in patients with clinical photosensitivity is higher than the frequency of ichthyosis in the global TTD population (Table IV) These results highlight the association between congenital ichthyosis and group A TTD with abnormal DNA repair.
Table IV. Frequency of Ichthyosis in Patients With Clinical Photosensitivity and in Global TTD Population
Total TTD population
N = 39
N = 122
Our main finding is a significantly higher frequency of neonatal ichthyosis in TFIIH-related TTD (group A) patients, as compared to the non-TFIIH related group (group B). Ichthyosis in group A has a mild course and looks like ichthyosis vulgaris. An aspect of collodion baby can be observed at birth in nearly a third of the patients. Among the environmental modifications, which could lead to the more severe cutaneous phenotype in the neonatal period, temperature might be considered. Cyclic hair loss has been described in association with fever in four patients belonging to group A [Kleijer et al., 1994]. This feature has been associated with the XPD p.Arg658Cys mutation which gives rise to a thermosensitive XPD protein. Elevation of temperature would be responsible for a worsening of DNA repair and transcription anomalies leading clinically to hair loss and increase of severity of ichthyosis [Vermeulen et al., 2001].
In TTD cells, abnormal TFIIH needs to be produced quickly enough to compensate its instability [Botta et al., 2002]. In differentiating TTD cells, de novo synthesis is insufficient and leads to accumulation of inactive factors and depression of basal transcription particularly for genes involved in terminal epidermal differentiation or neuronal myelination and pseudo-thalassemia. Mutations in epidermal differentiation genes (TGM1 (MIM190195), ALOXE3 (MIM607206), ALOX12B (MIM603741), ABCA12 (MIM607800), ichthyin (MIM609383), loricrin (MIM152445)) involved in the congenital ichthyoses manifest also commonly by a collodion baby phenotype. It could thus be speculated that the clinical and histological aspects of ichthyosis vulgaris observed in TTD patients could result of the defective expression of epidermal differentiation complex proteins, most of which are located in 1q21 including filaggrin and SPRR2. Skin biopsies of ichthyosis performed in group A patients had similarities with ichthyosis vulgaris, with diminished keratohyalin expression, a marker of profilaggrin (Fig. 2). Furthermore, a diminution of SPRR2 expression has been found in the TTD mouse model homozygous for the XpdR722W allele [De Boer et al., 1998]. In vitro reconstruction of TTD human epidermis could allow to study the modification of the expression of proteins involved in keratinocyte differentiation and their dependence on temperature.
Instability and dysfunction of TFIIH with mutated sub-units could account for TTD findings [Schultz et al., 2000; Dubaele et al., 2003]. It has been shown that TFIIH activates transcription by phosphorylation of nuclear receptors such as RARα or thyroid hormone receptor through its CDK subunit [Htun et al., 1996; Liu et al., 2005]. Mutations in C terminal domain of XPD are directly responsible for a decrease of phosphorylation of nuclear receptors and expression of targeted genes [Keriel et al., 2002]. In particular, it has been shown that the phosphorylation of the N-terminal domain of the γ subunit of the RAR by the CDK7 component of TFIIH leads to receptor activation through modulation of its interaction with a coregulator (vinexin β) [Bour et al., 2007].
Hypogonadism, which was initially described in Amish and Moroccan patients who belong to group B, is also found in group A. The difference of frequency between the two groups was significant but more patients with TTDN1 mutations have to be studied to make definitive conclusions. TTDN1 mutants responsible for group II TTD could be involved in the maturation of spermatozoid sulfur-rich proteins. In the drosophila, spermatogenesis is sensitive to β2-tubulin level, a protein of the tubula. XPD mutations in drosophila affect β2-tubulin leading to sterility in males. This mechanism could explain gonadal immaturity in TTD group A patients [Raff et al., 1982].
Osseous anomalies were found at a similar frequency in groups I and II, suggesting that abnormal function of both TFIIH or of TTDN1 could affect bone formation. The TTD-Xpd mouse model provides a good clinical reproduction of the disease for bony anomalies, which indicates a potent modulation of bone mineralization by abnormal TFIIH [De Boer et al., 1998, 2002].
Unlike xeroderma pigmentosum, TTD is not a cancer prone-disease. The group I cells are unable to repair the major cyclobutane pyrimidine dimers (CPD) induced by solar UV. However, the TTD cells mutated on the 5′ part of the XPD gene are also defective in the repair of the second type of UV-induced DNA lesions (pyrimidine 6-4 pyrimidone), while the cells mutated on the 3′ part of the same gene are proficient in this repair [Chiganças et al., 2008]. Nevertheless, none of these TTD patients are cancer-prone and therefore the cancer-free phenotype in TTD should not be directly related to a DNA repair defect [Nishiwaki et al., 2004]. On the other hand, differences in cellular catalase activity between TTD and XP indicate that UV light, directly or indirectly, together with defective oxidative metabolism may increase the initiation and/or the progression steps in the XP environment compared to TTD [Vuillaume et al., 1992]. It has been shown on normal and XP human reconstructed epidermis that catalase overexpression had a protective effect against deleterious effects of UV irradiation [Rezvani et al., 2007, 2008]. This may partly explain the differences in skin tumor proneness between group A TTD and XP.
In conclusion, TTD regroups recessively inherited affections, which have in common a specific hair dysplasia. Molecular studies suggest to classify TTD in three genetic groups. Mutations in the three genes encoding TFIIH subunits (XPD, XPB, p8) are responsible for the in vitro photosensitive form (group I). The non-photosensitive group is genetically heterogeneous including TTDN1 mutated patients (group II) and a third group without known molecular basis. Our phenotype/genotype correlation study showed a highly significant association between ichthyosis and group I, and the collodion baby phenotype gives an early diagnostic orientation for this group, without being completely specific. This classification is presented with its genetic and clinical correlations in Table V.