Key words
antibodies to thyroperoxidase - autoantibodies - autoimmunity - autoimmune thyroiditis
- Gravesʼ disease - Hashimoto’s disease
Introduction
Autoimmune thyroid diseases (AITD) range from the hyperthyroidism of Gravesʼ disease
(GD) to destructive Hashimotoʼs thyroiditis (HT), which leads to hypothyroidism. Most
patients with AITD are positive for autoantibodies directed against thyroid peroxidase
(TPO) and usually also for thyroglobulin (Tg). Autoantibodies against TPO (TPOAbs)
and TPO-specific T cells play important roles in the autoimmune destruction of thyrocytes
[1]
[2]
[3]
[4]. TPO is also the key enzyme involved in the generation of thyroid hormones in the
apical surface of thyroid epithelial cells [5]
[6]. In this short commentary, we present an update on TPO, with latest findings reported
over the past 5 years.
TPO gene and protein
Human thyroid peroxidase is encoded by TPO gene located on chromosome 2p25 and its expression is regulated by multiple factors
[6]. Inactivating mutations in TPO gene lead to congenital hypothyroidism resulting from thyroid dyshormonogenesis [7]
[8]
[9]. TPO mRNA composed of 17 exons is most abundant in human thyrocytes and encodes the 933-amino
acid TPO-1 protein. Due to alternative splicing, multiple transcripts without at least
one exon are also generated [6]. TPO-1 is a dimeric glycoprotein with covalently linked heme, where the large N-terminal
extracellular region (ectodomain) projects into the follicular lumen. The protein
also has a short transmembrane domain and intracellular C-terminal region. The TPO
ectodomain shows areas with homology to other proteins: myeloperoxidase (MPO-like
domain; residues 142–738), complement control protein (CCP-like domain; residues 739–795),
and epidermal growth factor (EGF-like domain; residues 796–846) [4]
[6]
[10]. Presently, the full tertiary structure of TPO remains unknown, however, high sequence
identity to proteins of known structure has provided some insights into the three-dimensional
organization of TPO by homology modeling approaches [11]. Recently, a model of full-length, membrane-bound dimeric TPO has been generated
[4]
[10]. Molecular models were used extensively to predict the structure of epitopes recognized
by TPO-specific autoantibodies, but a complete understanding eagerly awaits high resolution
structure determination.
Characteristics of TPOAbs
Almost all HT patients and nearly 75% of individuals with diagnosed GD have detectable
TPOAbs [1]. These autoantibodies are also frequently present in euthyroid subjects, particularly
women, even within the normal range for thyrotropin (TSH). TPOAbs level is one of
the predictive factors for conversion from euthyroidism to thyroid dysfunction [12]. Recently published data on 10-year observation of TPOAbs variations in a population-based
prospective study in Iran showed an increasing trend in TPOAbs levels over time accompanied
by the rise in hypothyroidism incidence [13]. Another recent Dutch population-based prospective cohort study in the elderly (aged
85 and older) also confirmed an increased risk of subclinical and overt hypothyroidism
in individuals with elevated TPOAbs [14]. Moreover, TPOAbs-positivity in oldest old community is associated with higher TSH
level and a decreased 10-year mortality risk [14]. Legakis et al. have shown that in the first months of life, the variability of
TPOAbs titers in neonates and infants is higher in boys than in girls [15]. The cut-off between TPOAb-positive and negative sera continue to be debated [16]
[17]
[18]. TPOAbs values may differ significantly between different immunoassays due to TPO
antigen preparations (native or recombinant antigen purified preparations) used to
coat solid phase ELISA plates [19]. TPOAbs in patientʼs sera are polyclonal and belong predominantly to the IgG class.
They usually have high affinity for TPO and preferentially bind to conformationally
intact protein [1]. Interestingly, a recently identified type of natural bispecific antibody against
TPO and Tg may play a protective role in the pathogenesis of HT [20]. Extensive studies conducted by various groups on genetic loci associated with the
presence of TPOAbs revealed their localization, for example, in the HLA region and inside or near genes encoding TPO, TSH receptor (TSHR), and Tg [21]
[22]
[23]
[24]
[25]
[26]
[27].
Epitope organization of TPO
Human TPOAbs recognize discontinuous determinants on TPO called immunodominant region
A (IDR-A) and B (IDR-B) [1]
[4]
[6]. Numerous studies have restricted IDR-A and -B to MPO-like domain and to a lesser
extent the CCP-like domain [10]. A number of contact residues that constitute IDR-A and IDR-B have been identified:
225, 353–363, 377–386, 597–604, 611–618, 620, 624, 627, 630, 646, 707, 713–720, and
766–775 [6]
[28]. Spatial arrangement of epitopes in the context of oligomeric state, domain architecture
and positioning in the membrane suggests that interaction of TPO with autoantibodies
may require significant changes in the overall tertiary structure of the antigen [4]
[10]. The distribution of IDR-A and -B-specific TPOAbs is similar in HT and GD patients,
where the TPO epitope pattern is inherited in families and stable over time [6].
Role of TPO and TPOAbs in thyroid dysfunctions
A high incidence of TPOAbs in AITD patientsʼ sera justified numerous attempts to understand
their role in thyroid pathology. It is postulated that TPOAbs can damage thyrocytes
by antibody-dependent cell cytotoxity, activation of complement cascade, and cell
damage [29]. TPOAbs-positivity has been recently found to be a main risk factor for developing
oxidative stress in euthyroid individuals with HT [30]. On the other hand, dietary phenolic antioxidants can interact with TPO leading
to its enzymatic activity inhibition [31]. The putative factors contributing to the breaking in the self-tolerance to TPO
and other thyroid antigens have been summarized [2]. Since then, Liu et al. has reported that serum TPOAbs titers are negatively correlated
with serum selenium concentrations in GD patients [32]. More recently, Wang et al. published results of a randomized, placebo-controlled
study, in which selenium supplementation decreased TPOAbs titers in serum positive
patients with autoimmune thyroiditis [33]. Individual patients exhibited different decreases in TPOAbs levels that may be
associated with genetic background [33]. Animal model studies confirmed that low dietary selenium potentiates the development
of TPOAbs and TgAbs in female mice; on the other hand, the authors did not observe
any reduction in thyroid antibodies titers due to higher dietary selenium administration
[34]. Numerous studies have also shown that the prevalence of TPOAb-positivity in individuals
is correlated with increased dietary iodine intake [35]
[36]
[37]
[38].
Clinical utility of TPOAbs testing
As mentioned above, TPOAbs are a useful marker in AITD diagnosis (especially in hypothyroid
patients), whereas in euthyroid patients TPOAbs positivity may increase the risk of
future thyroid disorders [1]
[39]. In recent years, new findings regarding the connection between TPOAbs and Gravesʼ
orbitopathy (GO) development in children have been revealed. Some studies did not
find a significant correlation between them [40], whereas others show a positive association [41]
[42]. In addition, studies on TPOAbs as risk factors of GO development in adult individuals
with GD can provide conflicting data [43]
[44]. Recently, much interest has been attracted to the role of thyroid autoimmunity
in fertility [45]
[46]
[47]. TPOAb positivity, even in euthyroid women, may have a connection with subfertility
and increased rates of pregnancy complications, including preterm delivery and miscarriage
[48]. Therefore, the monitoring of TPOAbs indices is recommended for thyroid screening
during pregnancy in the latest guidelines [49]. A recent prospective birth cohort study reported that TPOAbs-positivity during
early pregnancy is associated with lower child IQ in Dutch, but not in the United
Kingdom population [50]. Iodine status (lower in the latter) may be one of the reasons for these differences
[50]. TPOAbs are able to cross the human placenta and their titers in cord blood at the
moment of birth are similar to third-trimester maternal concentrations [51]. Nevertheless this passage is likely not associated with fetal thyroid dysfunction
[49]. The phenomenon of TPOAbs- and TgAbs-attenuated thyroidal response to human chronic
gonadotropin (hCG) during the first half of pregnancy has been recently reported [52]
[53]. In particular, it is known that serum hCG, due to its high homology to TSH, can
bind the TSH receptor on thyrocytes and stimulate secretion of thyroid hormones [54]. Detection of TPOAbs during pregnancy can predict the development of postpartum
thyroiditis [49]. Increased interest has been focused on the association between thyroid function
and metabolic disorders in euthyroid individuals [55]
[56]. For example, Liu et al. have recently shown that TPOAbs level is associated with
cardiometabolic risk factors in non-obese euthyroid individuals [56]. Another widely discussed issue is the connection between thyroid antibodies and
cancer [57]
[58]
[59]. Some authors reported positive association between anti-thyroid autoantibodies
(TgAbs and/or TPOAbs) titers and papillary thyroid cancer (PTC) risk and severity
[60]
[61], however, others question the role of thyroid autoimmunity in thyroid cancer [62]. The large majority of studies reported higher frequency of thyroid antibodies,
especially TPOAbs, in breast cancer patients in comparison with healthy controls [59]. Several reports showed a more favorable outcome of breast cancer in TPOAbs-positive
patients than in TPO-negative patients, nevertheless the protective role of TPO-driven
autoimmunity continues to be debated [59]
[63]. One hypothesis suggested shared antigen(s) between thyroid and breast tissues,
which trigger immune reaction in thyroid autoimmunity and breast cancer. Recent findings
showing TPO expression in tumoral breast and adjacent breast peri-tumoral tissues
supported this assumption [64]
[65]. Biochemical properties of thyroid and breast TPO are similar, with only some differences
[66]. Importantly, breast TPO was effectively recognized by a broad panel of conformation-sensitive
mAbs and human autoantibodies from AITD patients [65]
[66]. Patients with non-thyroid autoimmune disorders (e. g., type 1 diabetes) or congenital
disorders (e. g., Downʼs and Turnerʼs syndromes) predispose individuals to TPOAbs
development [1]
[67]
[68]
[69].
Conclusions
Despite significant progress in TPO research, there are still many questions left
unanswered. An atomic resolution structure of TPO remains elusive, and thus a detailed
molecular understanding of its antigenicity. Moreover, the mechanism of breaking self-tolerance
to TPO as well as the role of TPO antibody-positivity in various pathologies (e. g.,
breast and thyroid cancer) remain to be fully elucidated.
