2.1. Transglutaminase 2
By the 1960s, it was clear that, in addition to antibodies against the immunostimulatory gluten proteins of wheat and related cereals, patients with celiac disease express elevated autoantibody responses to the connective tissue surrounding smooth muscle fibers [
19,
20]. These antibodies became known as anti-endomysial and anti-reticulin antibodies [
21,
22,
23]. It was not until 1997, when Dieterich et al. identified TG2 as the major endomysial target autoantigen [
24]. Later research has shown that the target of the so-called anti-reticulin antibodies was also TG2 [
25].
TG2 is a member of the structurally and functionally related group of transglutaminase proteins that catalyze the modification of proteins by introducing covalent bonds between amine groups (such as a lysine) and γ-carboxamide groups of peptide-bound glutamines [
26]. TG2, the first transglutaminase discovered, is unique in some aspects, including that it is a ubiquitously expressed enzyme expressed in various tissues and cell types, and in various locations inside the cell and at the cell surface [
26]. Under certain conditions, TG2 may react with H
2O in preference over an amine, converting a specific glutamine to glutamate via deamidation [
27]. Intracellular TG2 is involved in signaling events that support cell survival in response to wounding, hypoxia, and oxidative stress [
27]. Extracellular TG2 has a role in the regulation of the cytoskeleton by crosslinking extracellular matrix proteins such as fibronectin and integrins, and is believed to function in cell adhesion, matrix assembly, and cell motility [
28].
TG2 plays a central role in the initiation of immune reactivity towards dietary gluten in the context of celiac disease [
29,
30], which also involves the celiac disease susceptibility genes, HLA-DQ2 and -DQ8 [
31]. TG2 can effectively deamidate specific glutamine residues of gluten sequences that may have crossed the epithelial barrier and found access to the lamina propria. Antigen presenting cells expressing the HLA-DQ2 and HLA-DQ8 molecules have an increased affinity for these deamidated peptides [
32]. Subsequent binding of the generated immunogenic peptides to the HLA molecules results in peptide complexes that can activate host gluten specific CD4
+ T cells in the lamina propria. Activation of these T cells is accompanied by the production of a number of cytokines that can in turn promote inflammation and villous damage in the small intestine through the release of metalloproteinases by fibroblasts and inflammatory cells, as well as providing help to activate gluten-specific B cell responses [
30,
33].
Gluten-specific CD4
+ T cells are speculated to stimulate B cell production of not only anti-gluten antibody, but also anti-TG2 antibody. In the absence of any TG2-specific T cells being identified, the anti-TG2 antibody response is believed to be driven by a process referred to as intermolecular help, similar to the hapten-carrier system. Accordingly, gluten-specific CD4
+ T cells are thought to provide help to TG2-specific B cells when TG2–gluten complexes are formed [
34]. Such a gluten-specific T cell-driven mechanism would lead to an anti-TG2 immune response without the requirement for TG2-specific T cells. With repeated exposure to TG2–gluten complexes, affinity maturation towards the TG2 antigen can potentially generate specific high-affinity autoantibody reactivity [
35,
36,
37]. Anti-TG2 autoantibodies are thus gluten-dependent, and the antibody titer decreases rapidly after the elimination of gluten from diet [
18,
38]. IgA anti-TG2 autoantibodies have high specificity (>90%) and sensitivity (>95%) in celiac disease, currently serving as a particularly useful aid in diagnosis [
18,
39].
Whether antibodies to TG2 can play a clear role in disease pathogenesis in humans has not been definitively proven. Anti-TG2 antibodies bind to several epitopes of TG2, including the enzymatic core of the protein, and can thus interfere with TG2 bioactivity [
40]. As TG2 is involved in epithelial cell differentiation through activation of transforming growth factor β [
41], anti-TG2 autoantibodies have been shown to reduce epithelial cell differentiation, increase epithelial cell permeability in an intestinal cell line, and induce monocyte activation upon binding to Toll-like receptor 4 [
42,
43,
44]. Data from in vitro studies indicate that anti-TG2 antibodies detected along the villous and crypt basement membranes in the jejunum from celiac disease patients may take part in the intestinal damage, particularly the remodeling of the small bowel mucosal architecture and the development of villous atrophy as well as crypt hyperplasia [
43,
45,
46,
47].
Because TG2 is the most widely expressed member of the transglutaminase family of proteins in the body, being present in almost all cell types, and participates in various biological reactions, the autoantibodies have the potential to negatively affect the activity of the enzyme and its biological role in tissues outside of the gastrointestinal tract as well [
48]. The presence of IgA deposits colocalizing with TG2 in the liver, lymph nodes, muscle, thyroid, bone, and brain indicates that the circulating autoantibodies originating from the small intestine can access the autoantigen throughout the body and may potentially exert certain pathogenic effects [
42,
49,
50,
51]. Although the significance of anti-TG2 antibody binding to thyroid tissue or bone is not clear, both thyroid dysfunction and reduced bone density are common in celiac disease, raising the possibility that the antibodies may affect target organ function [
49,
50,
52]. Anti-TG2 autoantibodies found within the muscular layer of brain vessels have also been speculated to cause disruption of the blood–brain barrier, which may further expose the central nervous system to other autoantibodies and potential toxins [
51]. In mice, the injection of anti-TG2 antibodies in the lateral ventricle of the brain has been shown to cause deficits in motor coordination [
53]. The data suggest that once exposed to the central nervous system, anti-TG2 antibodies may play a role in inducing neurologic deficits. Anti-TG2 antibodies isolated from celiac disease patients also have the potential to cross-react with other members of the transglutaminase family of enzymes due to some level of sequence homology, suggesting that such autoantibodies may additionally affect the activity of other transglutaminases [
53]. Taken together, the data are suggestive of a pathogenic role for anti-TG2 autoantibodies in some of the extraintestinal manifestations of celiac disease.
2.2. Transglutaminase 3
The skin manifestation of celiac disease, known as dermatitis herpetiformis, was first described by Louis Adolphus Duhring in 1884 as an itchy, blistering, skin disease [
54]. Dermatitis herpetiformis is characterized by the deposition of pathognomonic granular IgA in the dermal papillae, sometimes without significant gastrointestinal symptoms, and shares the same HLA associations with celiac disease [
5]. In 2002, epidermal transglutaminase (eTG), also known as transglutaminase 3 (TG3), was identified as the main autoantigen target in skin IgA deposits in dermatitis herpetiformis [
55]. In addition to increased anti-TG2 antibodies, dermatitis herpetiformis patients also present with elevated levels of antibody directed at TG3 [
55,
56].
TG3 is mainly expressed in the cornified layer of the epidermis and has been shown to play an important role in epidermal keratinization and in the formation of cornified envelope, which is essential for the maintenance for skin homeostasis [
57]. Serum levels of anti-TG2 and anti-TG3 antibodies appear to correlate in celiac disease patients without skin manifestation, but not in patients with dermatitis herpetiformis, suggesting there is antibody reactivity to specific non-cross-reactive epitopes of TG3 in dermatitis herpetiformis [
55,
58]. IgA autoantibodies against TG3 have been reported to be detected in as much as 95% of dermatitis herpetiformis patients, substantially more than those against TG2 (79% of patients) [
59]. In addition, detection of IgA antibody to TG3 has been found to efficiently distinguish untreated dermatitis herpetiformis from other dermatological itchy diseases and to be highly sensitive to a gluten-free diet [
60]. IgA anti-TG3 antibody has been proposed as a useful diagnostic marker for dermatitis herpetiformis in both pediatric and adult patients [
59,
60,
61].
The production of anti-TG3 antibodies may begin as a result of cross-reactivity of anti-TG2 IgA antibodies with TG3, which is released from epidermal keratinocytes and can diffuse through the basement membrane in regions of trauma [
62]. Prolonged gluten immune stimulation may allow for epitope spreading and further maturation of these antibodies, resulting in the development of high affinity anti-TG3 antibodies [
55,
62,
63]. Disappearance of anti-TG3 IgA antibody in response to dietary exclusion of gluten is slow and may take longer than for antibody response to TG2, suggesting that mechanisms other than homology between TG2 and TG3 might trigger the production of anti-TG3 antibodies [
64,
65,
66].
Deposition of IgA antibodies in dermatitis herpetiformis is believed to play a role in the infiltration of neutrophils into the papillary dermis and in the formation of basement membrane zone vesicles in the lamina lucida [
55]. TG3 in the papillary dermis has been found to overlap with the deposits of IgA antibodies in dermatitis herpetiformis patients, implying that TG3 is bound by the IgA autoantibodies [
55,
63]. It is hypothesized that active TG3 may cross-link anti-TG3 antibodies to certain dermal structural elements, leading to the dermal deposition of anti-TG3 IgA, which can in turn invoke skin pathology such as the associated blisters and papules [
55]. However, anti-TG3 IgA deposits have also been found in uninvolved skin in affected patients, in areas away from lesions, suggesting that factors beyond these immune complexes may be necessary for lesion formation [
62,
63].
2.3. Transglutaminase 6
In addition to the well-characterized gastrointestinal and skin manifestations, a number of studies have reported on various other symptoms associated with celiac disease. Neurologic deficits, including peripheral neuropathy and cerebellar ataxia, are among the most common extraintestinal symptoms reported in conjunction with celiac disease [
67,
68]. Furthermore, elevated levels of anti-gliadin antibody have been associated with idiopathic neuropathy and idiopathic ataxia, even in the apparent absence of the characteristic mucosal pathology [
69,
70,
71,
72]. The terms “gluten ataxia” and “gluten neuropathy” have been used to describe these conditions, although the significance of the anti-gliadin antibodies in the absence of biopsy-proven intestinal damage has been debated [
70,
72,
73].
Among these, idiopathic or sporadic ataxia associated with anti-gliadin antibodies has been the best studied in terms of understanding its frequency in different populations of ataxia patients and its potential etiology and pathogenic mechanism. A recent meta-analysis of several studies further validates the presence of significantly increased levels of antibody to gliadin among patients with non-hereditary ataxia [
74]. There have been suggestions that gluten ataxia would fit better within the spectrum of non-celiac wheat/gluten sensitivity (NCWS) rather than celiac disease, based on serologic, histologic, and genetic markers [
75,
76]. In 2008, a novel neuronal transglutaminase, TG6, was reported as a target autoantigen in patients with gluten ataxia [
77]. Antibodies to TG6 were later also detected in patients with gluten neuropathy [
78]. However, the specificity of anti-TG6 antibodies in gluten ataxia and gluten neuropathy needs further investigation, as other studies have found such antibodies in patients with other conditions and in those without neurologic symptoms as well [
77,
78,
79,
80,
81].
TG6 is predominantly expressed in a subset of neurons and plays a role in neurogenesis, particularly in the context of nervous system development and motor function [
82]. TG6 is encoded on the same chromosome (20q11–12) as TG2 in humans [
83]. Similarly to TG2, when TG6 is incubated with gluten peptides, it can both deamidate and transamidate glutamine residues, and there is a large degree of overlap in glutamine donor substrates of TG6 and TG2 [
82]. TG6 can also form the previously mentioned hapten-carrier complexes with gluten, but to a lesser extent when compared with TG2 [
84]. Therefore, it is conceivable that, in the event of blood–brain or blood–nerve barrier disruption, TG6 may become exposed to gluten-derived antigens. As such, gluten-specific CD4
+ T cells may be able to provide help to TG6-specific B cells, leading to the production of anti-TG6 autoantibodies in a similar fashion to anti-TG2 antibodies.
The relationship between gluten intake and the development of anti-TG6 antibodies has been examined. Lindfors et al. did not observe a decrease in anti-TG6 IgA antibodies in celiac disease patients on a gluten-free diet, suggesting a lack of gluten-dependency of TG6 autoantibodies [
81]. However, a more recent study on a pediatric cohort of celiac disease patients found anti-TG6 antibody reactivity to correlate with the duration of gluten exposure, and to decline in response to the introduction of a gluten-free diet [
79].
Similar to what has been described for anti-TG2 autoantibodies, anti-TG6 antibodies have the potential to disrupt TG6’s biological functions. In mice, the injection of celiac disease patient-derived TG2 antibody that can cross-react with TG6 into mouse brain has been shown to cause deficits in motor coordination [
53]. Recently, missense mutations in the of transglutaminase 6 gene have been identified in families of Chinese patients with spinocerebellar ataxia type 35 (SCA35) [
85]. However, a causative link between neurological manifestations and autoantibodies to TG6 remains unclear [
86]. Future studies on anti-TG6 antibodies can help in further clarifying their diagnostic and pathogenic potential in the context of neurologic and other manifestations in celiac disease and gluten sensitivity.