Figure 1.11. Model for mechanism of global genome nucleotide-excision repair and
transcription-coupled repair (Hoeijmakers et al., 2001).
1.3.6. hHR23 (RAD23) Genes Human hHR23A and hHR23B are homologs of Saccharomyces cerevisiae Rad23
and Saccharomyces pombe Rhp23 genes. hHR23A and hHR23B have been mapped to
chromosome 19p13.2 and 9q31.2 encoding 363 and 409 amino acid proteins, respectively.
An amino acid sequence comparison of
the Rad23, Rhp23, hHR23A and hHR23B showed
of at least four distinct domains that are well conserved among
(van der Spek et al, 1996). First, they contain a novel domain near the N-terminus referred
to as UBL (ubiquitin-like) domain which is important for DNA repair. This domain can
to the 26S proteasome, and the removal of it from the yeast
interaction with the proteasome and was associated
with increased sensitivity to UV light
(Watkins et al, 1993;
Schauber et al, 1998). The second and fourth domains from the
terminus are Ub-associated (UBA) domains, suggesting
the involvement of hHR23 in
certain pathways of Ub metabolism
(Hofmann and Bucher, 1996). The third domain, the
phosphoprotein (STI1), has been found to be responsible
for binding the XP-
C protein (Masutani et al, 1997) (Figure 1.12).
Figure 1.12. Schematic representations of conserved domains in hHR23A (a) and hHR23B
XPC-hHR23 complex is the initial component for detecting DNA damage in GG-
NER pathway. Although both hHR23 proteins can bind to XPC at least in vitro (Sugasawa
et al ., 1997), most XPC is complexed with hHR23B in vivo, and only a minor fraction of
the complex contains hHR23A (Araki et al., 2001).
Studies on hHR23A/B knock-out (KO) mice have shown that hHR23A was likely
functionally redundant with hHR23B. hHR23B KO mice exhibit moderate UV sensitivity
and NER deficiency, whereas hHR23A KO mice do not exhibit any observable defects in
DNA repair activity. However, when both hHR23A and hHR23B were deleted, the mice
exhibited extremely severe phenotypes, impaired embryonic development and high rates of
intrauterine death. Surviving animals displayed retarded growth, male sterility, facial
dysmorphology and DNA repair defects (Ng et al., 2003). Since these phenotypes were not
observed in XPC KO mice, the hHR23A/B KO phenotype was not due to NER deficiency
suggesting that hHR23A/B proteins have additional cellular functions. These findings
indicate also that the hHR23A does not completely duplicate the function of hHR23B and
must have a function of its own. However, mHR23A and mHR23B appeared to have
redundant roles in NER. Over expression of hHR23A in the mHR23A/B double knock out
cells restored not only the steady-state level and stability of the XPC protein, but also
cellular NER activity to near wild-type levels (Okuda et al., 2004).
Studies on hHR23A/B knockdown (KD) cell lines have shown that while hHR23A
cells were not blocked in S phase after UVC irradiation, many hHR23B
hindered in getting out of S phase. This suggested the presence of unrepaired UVC-
induced DNA damage in hHR23B
cells. Therefore, hHR23B
cells seemed to behave
like XP cells. hHR23B
cells displayed a significant sensitivity to UVC, in contrast to
cells, which strongly tolerated UVC irradiation (Biard A, 2007). This also
suggested that hHR23A and hHR23B displayed diverse biological functions leading cells
to different outcomes.
Intriguingly, only a minority of hHR23B and hHR23A is bound to XPC, suggesting
that both proteins have additional functions (Sugasawa et al., 1996). hHR23 proteins are
players for multiple mechanisms including DNA repair and proteasome-mediated protein-
degradation and apoptosis (Kim et al., 2004; Glockzin et al., 2003).
hHR23 proteins connect the NER and ubiquitin/proteasome mediated protein
degradation pathways via UBA and UBL domains. The ubiquitin-proteasome pathway
plays a key regulatory role in a variety of cellular events, including the removal of
misfolded proteins, production of immunocompetent peptides, activation or repression of
transcription, and regulation of cell cycle progression (Schubert et al., 2000; Yamaguchi et al ., 2000). Proteins are ubiquitylated and consequently delivered to the 26S proteasome for
degradation. Ubiquitin receptor family can directly connect ubiquitylated proteins to the
proteasome via their UBA and UBL domains binding the ubiquitin and proteasome,
respectively. However, depending on the levels of UBL/UBA containing proteins they can
promote or inhibit the degradation of ubiquitylated substrates. The binding of UBL/UBA
containing proteins to the ubiquitylated substrates inhibits the polyubiquitination and
prevent proteolysis by proteasomes (Ortolan et al., 2000).
hHR23 proteins regulate the induction and stability of XPC via inhibiting the
proteolysis by proteasomes. In the absence of hHR23 proteins, XPC is highly unstable
since it is ubiquitylated and degraded by 26S proteasome. Under normal conditions, XPC-
hHR23 complex results in a significant reduction of XPC proteolysis and consequently in
increased steady-state levels of the protein complex. This correlates with proficient GG-
NER activity. The protecting role of hHR23 is performed via inhibition of
polyubiquitination (Schauber et al., 1998; Lommel et al., 2002; Ortolan et al., 2000). This
hypothesis was supported by the observation of presence of XPC-Ub conjugates and
increased XPC stability in proteasome inhibitor treated mHR23A/mHR23B KO cells (Ng
et al ., 2003).
In addition to NER, XPC–hHR23B complex is associated with the base excision
repair (BER). The XPC protein interacts physically and functionally with the thymine
DNA glycosylase, the enzyme that recognizes cyclobutane pyrimidine dimers at the initial
step of BER (Shimizu et al., 2003). 3-Methyladenine-DNA glycosylase, an initiator of
BER also interacts with hHR23A proteins, suggesting that these proteins are involved in
BER (Miao et al., 2000). Hsieh et al. reported that hHR23A
, but not hHR23B
were hypersensitive to the treatment of methylmethane sulfonate, a major substrate for
BER (Hsieh et al., 2005). This suggests that hHR23A might be a player for the two major
DNA repair pathways, BER and NER.
Moreover, multiple engagements between hHR23/Rad23 and cell cycle regulation
are present. (1) RAD23 has a partially redundant role with and binds to RPN10 in the
G2/M transition (Lambertson et al. 1999). (2) RAD23 is involved in spindle assembly and
S-phase checkpoints (Clarke et al., 2001). (3) RAD23, together with DSK2, has a role in
spindle pole duplication (Biggins et al., 1996). The link with spindle pole duplication was
recently strengthened by the discovery of the centrosome factor CEN2 as the third
component of the XPC/HR23 complex (Araki et al., 2001). (4) hHR23 proteins themselves
appear to be regulated in a cell-cycle-dependent manner with specific degradation during
S- phase (Kumar et al., 1999).
The damage-signaling tumor-suppressor protein p53 is partly regulated by hHR23A
and hHR23B. hHR23A and a minor amount of hHR23B form a complex with p53.
hHR23A and B proteins downregulated the transactivating activity of p53 via inhibition of
the CREB (cyclic AMP-responsive element binding) protein, which acts as a coactivator of
p53 transcription (Zhu et al., 2001). Overexpression of hHR23A and B proteins has led to
the accumulation of ubiquitinated p53 and blocked p53 proteasome degradation (Glockzin
et al ., 2003). These paradoxical effects of hHR23 proteins on p53 degradation are
consistent with suggestions that the effects of hHR23 on degradation are highly sensitive to
stoichiometric variation (Raasi and Pickart, 2003; Verma et al., 2004). Additionally,
hHR23A and B interact with mouse double minute 2 (MDM2) protein. MDM2 contacts
with 20S core particle of the proteasome and functions to antagonize the stabilizing
function of hHR23 toward p53, directly promoting p53 recognition and degradation by the
proteasome (Brignone et al., 2004; Sdek et al., 2005).
Kaur et al. (2007) has shown that hHR23B was required for genotoxic-specific
activation of p53 and apoptosis. After exposure with UV or chemical agents leading to
DNA damage, p53–Ub conjugates accumulate in chromatin and hHR23B is required for
induction and maintenance of these p53–Ub species. The knockdown of hHR23B blocks
p53 stabilization and resulted in significant reduction in apoptosis and increase in viability
when compared to hHR23B expressing cells. Robust XPC depletion had no impact on
genotoxin-induced apoptosis suggesting that the inhibition of apoptosis due to hHR23B
depletion could not be explained by a reduction in DNA repair efficiency. The hHR23B-
dependent accumulation of p53–Ub conjugates after DNA damage correlated with p53
stabilization and apoptosis. p53–Ub conjugates could contribute to transcription-dependent
functions of p53, which are required for downstream p53 activities (induction of target
genes p21 and bax) (Slee et al., 2004; Schuler and Green, 2005).
1.4. Congenital Hypothyroidism (CH)
1.4.1. The Thyroid Gland
All vertebrates possess a pair of thyroid glands, located in the anterior neck region. It
consists of two lobes: one on either side of the trachea, and a connecting portion called the
isthmus, giving the entire gland an H-shaped appearance. The gland varies in size with
sexual development, diet and age. The thyroid gland consists of a large number of round or
oval follicles surrounded by connective tissue and blood vessels. Each follicle is lined by a
cuboidal epithelial cell layer of one-cell thickness. The cavities of the follicles are filled in
with viscous protein material called colloid in which the thyroid hormone, thyroxine, is
Normal thyroid function is essential for development, growth and metabolic
homeostasis. The primary function of the thyroid is the formation, storage, and secretion of
thyroid hormones (Robbins et al., 1980; Kohn et al., 1993). Thyroid hormone formation
involves a coordinated series of steps controlled by hypothalamic-pituitary-thyroid axis
(Figure 1.13). The TRH, released from hypothalamus, results in secretion of TSH from the
anterior pituitary gland. The TSH stimulates the follicular cells (TFC), via its receptor
(TSHr), to release thyroid hormones into the circulation. Thyroid hormone production
involves concentrative iodide uptake by the sodium iodide transporter (NIS), as well as
iodination of thyroglobulin (TG) by the thyroid peroxidase (TPO) (Robbins et al., 1980;
Kohn et al., 1993).
TG is synthesized as a 12S molecule (330 kDa), post-translationally glycosylated and
transported to follicular lumen. The protein contains 70 tyrosine residues that are subject to
iodination by TPO. The iodinated TG stored in follicles are degraded by lysosomes to form
triiodothyronine (T3) and thyroxine (T4). The major thyroid hormone secreted by the
thyroid gland is thyroxine. To exert its effects, T4 is converted to triiodothyronine (T3) by
the removal of an iodine atom. This occurs mainly in the liver and in certain tissues where
T3 acts, such as in the brain (Robbins et al., 1980; Weiss et al., 1984; Ekholm, 1990; Kohn
et al ., 1993; Dai et al., 1996).
Figure 1.13. Thyroid hormone cascade.
1.4.2. Congenital Hypothyroidism
Any defects in thyroid morphogenesis and hormone synthesis result in Congenital
Hypothyroidism (CH), which is a relatively common congenital disorder affecting about
1:3000 to 1:4000 live births (Toublanc, 1992). In about 10 per cent of all cases, CH is the
consequence of defects in one of the steps of thyroid hormone synthesis, inborn errors of
metabolism referred to as dyshormonogenesis. A heterogeneous group of developmental
abnormalities, thyroid dysgenesis, accounts for about 85 per cent of all cases with CH
(Gillam and Kopp, 2001). These anomalies include thyroid agenesis, ectopic thyroid tissue,
cysts of the thyroglossal duct, and thyroid hypoplasia. In the vast majority of all cases,
thyroid dysgenesis is sporadic, but in about 2 per cent it is a familial disorder, an
observation supporting the possibility of a genetic etiology (Castanet et al., 2000). The
higher prevalence of thyroid dysgenesis in Hispanics and Caucasians in comparison to
Blacks, the predominance of thyroid dysgenesis in females, and the higher prevalence of
associated malformations also suggest the presence of genetic factors in the pathogenesis
of CH (Devos et al., 1999; Castanet et al., 2001). More recently, the data from knockout
mice have demonstrated the roles of several genes in thyroid organogenesis; thyroid
transcription factors (TTF-1 and TTF-2), paired box homoetic gene 8 (Pax8), and TSH
receptor (TSHR). Occasionally mutations in these genes have been reported in CH cases.
Thyroid transcription factor 2 (TTF-2, FKHL15, or Forkhead Box E1 FOXE1) is a
member of the forkhead/winged helix-domain protein family, many of which are key
regulators of embryonic development. TTF-2 regulates the transcription of target genes
such as TG and TPO by binding to specific regulatory DNA sequences in their promoters
via its forkhead DNA binding domain. Homozygous recessive mutations (p.S57N and
p.A65V) in TTF-2 result in a syndromic form of thyroid dysgenesis, Bamforth-Lazarus
syndrome. This phenotype includes thyroid agenesis, cleft palate, choanal atresia, bifid
epiglottis, and spiky hair (Bamforth et al., 1989; Clifton-Bligh et al., 1998; Castanet et al.,
2002). Mice homozygous for a disrupted TTF-2 gene die within 48 hours after birth and
are profoundly hypothyroid. They exhibit either small lingual thyroid or have complete
thyroid agenesis, and also have severe cleft palates. Hair defects could not be tested since
the mice die before hair formation (De Felice et al., 1998).
1.4.3. Forkhead Gene Family
Forkhead transcription factors are key regulators of embryogenesis and play
important roles in cell differentiation and development. The name derives from two
spiked-head structures in embryos of the Drosophila fork head mutant, which are defective
in formation of the anterior and posterior gut (Weigel et al., 1989). A 110-amino-acid
DNA binding domain was evolutionarily conserved between forkhead genes (Figure 1.14).
These genes have so far been found in animals, fungi and most of the metazoans. Among
the organisms for which the genome sequences are completed, or nearly so, there is indeed
a correlation between anatomical complexity and forkhead gene number: four in
Saccharomyces and Schizosaccharomyces, 15 in Caenorhabditis, 20 in Drosophila, and 39
in Homo sapiens (Carlsson and Mahlapuu, 2002). X-ray crystallography revealed that the
3D structure of a forkhead domain (FoxA3) resembled the shape of a butterfly and the term
“winged helix” was used to describe the structure, which has a helix–turn–helix core of
three α-helices, flanked by two loops, or “wings” (Figure 1.15) (Clark et al., 1993). The
term “Winged helix proteins” is often used synonymously with forkhead proteins. A large
proportion of the amino acids in the forkhead domain are invariant or highly conserved
(Figure 1.14). Forkhead proteins bind DNA as monomers in contrast to other helix–turn–
helix proteins. Hence, the binding sites, which typically span 15–17 bp, are asymmetrical
(Pierrou et al., 1994).
Figure 1.14. Alignment of the TTF2 forkhead DNA-binding domain with selected human FOX representatives. At the bottom, the consensus
line indicates; ‘*’ for identical or conserved residues in all sequences; ‘:’ for conserved substitutions; and ‘.’ for semi-conserved substitutions.
The positions of predicted ‘helix’ and ‘wing’ segments are indicated at the bottom of the panel (Romanelli et al., 2003).
Figure 1.15. Three-dimentional structure of the forkhead domain of FoxC2 (mouse) (van
Dongen et al., 2000).
1.4.4. Human TTF-2 Gene Human TTF-2 gene, also known as FKHL15 or FOXE1, has been mapped on
chromosome 9q22 (Chadwick et al., 1997). TTF-2 consists of a single exon encoding for a
protein (42 kDa) of 367 amino acids. The gene encodes a forkhead domain, two nuclear
localization signals (NLS), a polyalanine tract, and a transcription repression domain
The sequencing of the entire coding region revealed a highly polymorphic
polyalanine stretch of 11 to 17 residues, but the most frequent stretch was 14 residues long
(Figure 1.16) (Macchia et al., 1999; Hishinuma et al., 2001). Reduced polyalanine tract
length (11 and 12 residues) was found only in patients with thyroid dysgenesis (Hishinuma
et al ., 2001). Similarly, alteration of polyalanine stretch lengths has been found in a
homeobox-containing gene, HOX D13, consisting of 15 residues (22-25 residues were
reported in normal) in patients with the autosomal dominant disease, synpolydactyly
(Akarsu et al., 1996). Functional analysis was performed to reveal whether reduced
number of polyalanine tract of TTF-2 is responsible for thyroid dysgenesis. However, the
expression study showed that the transcriptional activities of TTF-2 with reduced
polyalanine-tract lengths were equal to that of TTF-2 with an unreduced polyalanine tract.
These results suggested that the polymorphism of the polyalanine tract of TTF2 could not
be a cause of the developmental defects of the human thyroid gland (Hishinuma et al.,
TTF-2 protein contains two short stretches of basic amino acids (RRRKR) at both
ends of the forkhead domain (Figure 1.16). Sequence alignments of representative human
FOX protein segments show that a stretch of basic amino acids is present in most of the
proteins at the C-terminal of DNA-binding domain whereas a basic stretch at the N-
terminal of the DNA binding domain is less conserved (Figure 1.14). Previous studies have
demonstrated that the basic stretches at the C-terminal of the forkhead domain are bona fide NLSs in four FOX proteins (Brownawell et al., 2001; Berry et al., 2002). Romanelli et al. have shown that both stretches are bona fide NLSs and required for nuclear import via
importin α-dependent pathway (Romanelli et al., 2003).
Analysis of the TTF2 amino acid sequence identifies, in addition to the FHD, a
second domain rich in alanine and proline residues that has been found in other
developmental DNA binding proteins responsible for transcriptional repression activity.
The analysis of the activity of the deletion mutants of TTF-2 allowed mapping the
repression domain in the region between amino acids 197 and 218 at the carboxyl terminus
of the protein (Figure 1.16) (Perrone et al., 2000). Comparison of the amino acid sequence
of TTF-2 repression domain with the sequences present in the database demonstrates that
this region shows 41 per cent of identity with HNF3γ amino terminal. HNF3γ is another
member of the forkhead family that plays an important role in the tissue-specific gene
expression program both in early development and in the adult (Ang et al., 1993).
TTF2 is a promoter-specific transcriptional repressor that displays both promoter and
transcriptional activation domain specifity. TTF2 can repress Pax8 activity on TPO
promoter, but not on NIS promoter. Additionally, TTF2 is able to repress the C-terminal
activation domain of TTF1, while it has no effect on the N-terminal domain (Perrone et al.,
2000). These observations suggest that it is able to recognize specific promoter architecture
and represses only a subset of the genes activated by TTF1 and Pax8, since all these factors
are involved not only in thyroid specific gene expression (Plachov et al., 1990; Zannini et al ., 1992), but also in thyroid development (Kimura et al., 1996; Macchia et al., 1998;
Mansouri et al., 1998).
Figure 1.16. Nucleotide sequence of the human TTF2 gene and deduced amino acid
sequence. The alanine stretch is underlined. The forkhead, two NLSs and transcription
repression domains were boxed with black, blue and red color, respectively (Macchia et al ., 1999).