REPORT ON THE WORKSHOP "TRANSGLUTAMINASE, PROTEIN CROSSLINKING AND COELIAC DISEASE"
Carlo Bergamini, Italy
European Science Foundation has recently sponsored a conference in Tampere, Finland, (on Sept. 14-15) to collect together basic and clinical scientists to discuss the role of Transglutaminases in the coeliac disease (CD), under the chairs of Murkku Maki and Laszlo Fesus. The meeting was attended by 14 transglutaminase researchers and 75 preclinical and clinical researchers active on CD. A few aspects of the original program could not be covered because the invited speakers from USA (Mie-Jae Im, Cleveland, and Peter Steinert, NIH) were unable to come, for the well-known difficulties in transoceanic flights after the New York and Washington plane attacks. Even in their absence the meeting was quite interesting for partecipants from both sides and it is favourable for starting collaborations in these merging fields.
The conference topics were ideally devided into several sessions, starting with basic physiopathogenesis and clinics of Coeliac Disease. Markku Maki (Tampere University) reviewed epidemiology of CD, which is one of the most frequent genetic syndromes, with an estimated incidence of 1 over 150, including severe typical and mild atypical cases. CD is very polymorphous with presentation patterns which include severe malabsorption, diarrhea and malnutrition, severe or mild anemia, growth disturbances, seizures and other neurologic syndromes, osteoporotic, reproductive and cutaneous lesions and aspecific abdominal complaints. This multifaced appearence makes timely diagnosis difficult, despite its importance for the regression of clinical symptoms, upon admission to gluten-free diet, which is inversely related to the length of exposure to the alimentary antigen. The development of suitable diagnostic laboratory tools, as the detection of serum antibodies against tissular transglutaminase, has greatly helped in detecting atypical cases, giving further chance for appropriate preventive treatment, as further underlined, in his presentation, by Detlef Schuppan (University of Nurnberg-Erlangen), who summarized data supporting the involvement of TGase as main autoantigen of the CD, and its role in processing alimentary antigens through deamidation and crosslinkage, with important consequences for CD autoimmunity and for patientsí general immune reactivity.
In the subsequent session dealing with basic knowledge on expression, structure and activities of transglutaminases, Daniel Aeschlimann (University of Cardiff) reviewed his own work on the expression of multiple transglutaminase isoenzymes in animals and even in the same tissue (with possibility of multiple compensation as appearent also from the presentation of Gerry Melino), the evolutionary changes and the mechanisms of regulated expression of the different isoforms. Rolf Hilgenfeld (IMB, Jena) reviewed the recent progresses in high resolution structure of transglutaminase taking in account the x-rays diffraction patterns of Factor XIII, fish Type 2 transglutaminase and type 3 transglutaminase, underlying their similarities and differences, the presence of unusual non proline cis-peptide bonds, the reactivities at the active site and the potential strategies to develop transglutamisase-directed selective drugs. Carlo Bergamini (University of Ferrara) described the mechanism of regulation of the transamidating activity, through ligand induced conformational changes and movement of protein domains, which apparently also affect the stability and immunoreactivity of tipe 2 transglutaminase (it is now clear that CD patients develop at least two different types of antibodies directed against domain 1 and domain 4).
The following topics was related to functional properties of tissular transglutaminases in vitro and in vivo taking in account interaction of the enzyme with extracellular matrix components, the effects of disruption of its expression in vivo in transgenic mice models and the distribution of transglutaminase and antitransglutaminase antibodies in relation to atypical CD as the so-called dermatitis herpetiformis. In his presentation, Martin Griffin (Nottingham Trent University) focused on the subcellular distribution of tissular transglutaminase, providing data suggestive of the location of a significant fraction of cell transglutaminase activity, as a membrane exoenzyme, active in the interaction with and in the post-translational modification of EM proteins. These events are related to the distribution of fibronectin binding sites in the plasma membrane, with a colocalization of both proteins, are subjected to control by matrix metallo proteinases and to shear-stress following mechanical damage and are relevant for the regulation of cell motility. The effects of the experimental disruption of the expression of tissular transglutaminase in the mouse was examined in detail by Gerry Melino (University of Rome TorVergata) who had developed experimental transgenic animal models. Unexpectedly at the light of the massive in vitro data of crucial physiologic roles of the enzyme, the supression of the expression of tissular transglutaminase in vivo resulted into apparently normal phenotype, fertility and viability in the progeny. The only detectable anomaly was related to a diabetic diatesis as manifested by abnormal response to oral glucose loads. Apparently, suppression of the enzyme expression in vivo was largely compensated by overexpression of other transglutaminase isoenzymes, in particular the newly characterized type 5 transglutaminase, which mask most possible phenotypic defects. Miklos Sardi (University of Cologne) examined the effects and the accumulation of circulating IgA antibodies on the epidermal transglutaminases in vivo ad in vitro in patients affected by dermatitis herpetiformis (HD) which is also a gluten induced pathology, largely confined to the skin, without major involvement of the intestinal wall. The experimental evidence is that these patients develop antibodies against both type 2 and type 3 TGase in response to gliadin challenge, but that only antibodies against type 3 Tgase are mantained, for unknown reasons, during the disease progression.
The biology of the intestinal mucosa, including its regenerating potentials and the response to alimentary challenge represent the ideal topics of the subsequnet discussions. Patricia Simon-Assmann (INSERM U381, Strasbourg) analyzed structural biology of the basal membrane of the intestinal cell wall indicating how laminins direct the normal and pathologic development of the tissue. Laminis 1, 5, 8 and 10 seem to play pivotal role in the normal BM embriogenic development. They are variously secreted by the endodermal epithelial and by the mesenchimal cells, they assemble with each other on the nidogen receptors and are further polymerized by secreted TGase. Experimental disruption of this normal assembly either means of specific antibodies (e.g. against laminin 1) or by an antisense approach, is pathogenetic for CD disease but also for other inflammatory and neoplastic intestinal diseases. Thomas T Mac Donald (University of Southampton) discussed control of matrix metalloproteinases during growth and regeneration of intestinal mucosa. In inflammatory intestinal diseases, the intestinal wall itself, and particularly the lamina propria connective tissue, is haevily populated by inflammatory cells, including T-cells, neutrophiles, macrophages, eosinophiles and mastcells, which secrete cytokines and serine proteinases, altering normal cell-cell interactions. T-cell related inflammatory diseases are characterized by increased TNF secretion, leading to mucosal remodeling, largely through proteinases secretion. On line with this perspective, treatment with proteinases inhibitors frequnetly amelioretaes the experimental clinical picture. Marco Londei (Kennedy Institute of Rheumatology, London) described briefly his attempts to study induction of tissular transglutaminase by gliadin peptides in cultured explants of intestinal mucosa from CD and control subjects. These studies are still at a very preliminary stage although evidence was collected that primary cultures of CD and control mucosa display differences in transglutaminase expression upon gliadin challenge.
Discussion further moved to analysis of structure and toxicity of gluten proteins and to the genetic and immunologic basis of the disease. To assess primary toxicity of gliadin derived peptides, Herbert Wieser (Institute of Food Chemistry, Garsching, Germany) described the chemical structure of storage prolamin proteins of wheat, rye and barley (which are harmful to CD patients), while those of other cereals (notably rice and maize) are not toxic. The gliadins are alcohol-soluble prolamins (rich in proline and glutamines) present in forms of different molecular weight (high, medium and low, H, M and L). Peptides derived from the N-terminal domains ofa and g gliadin are particularly able to activate in vitro resident intestinal T-cells of CD patients, particularly if partially deamidated at glutamine residues by mild acid treatment. As further pointed out by Frits Koning (University of Leiden) the deamidation of glutamine in gliadin peptides in sensitive (HLA-DQ2 or DQ8) subjects can be carried out by intestinal transglutaminase and in important in breaking oral tolerance to gliadin proteins. His data further demonstrate that both not deamidated as well as deamidated peptides might be toxic and that additional peptides of the protein, not specific for eliciting initial response, might be involved in further progression and mantainence of the disease. As a consequence there is a different spectrum of response to gliadin peptides by T-cell repertoire isolated rom pediatric and adult patients. Ludvig M Sollid (University of Oslo) further contributed to this issue arguing that the known association of CD with HLA genotypes, as reported above, depends on the deamidation of specific glutamine residues to yield negatively charged peptides which form complexes with the DQ2 and DQ8 binding sites, to form the T-cells activatory epitopes. Evidence was presented that these regional specific deamidation were carried out in an enzyme dependent reaction, by intestinal transglutaminase. This issue was largely debated at the light of the balance between the hydrolytic and the crosslinking activity of the enzyme, the availability of additional amine substrates and the increased activity of transglutaminase in intestinal biopsis of CD than control patients.
The in situ features of the intestinal wall and the immune response to transglutaminase in CD were developed by several speakers. Carla Esposito (University of Salerno) proved the distribution of transglutaminase protein, substrates and accumulation of products in untreated celiac patients by a combination of immunocytochemical, semiquantitative PCR and in vitro activity approaches. Her results are consistent with an increased expression of tTGase in particular in the subepithelial layers of the intestinal wall, suggesting that the basal membrane remodeling activity of the protein is important in this respect. Hans Sjostrom (University of Copenhagen) further commented on the different distribution of transglutaminase in the subepithelial intestinal layers in adult and pediatric patients, leading to the conclusion that it mainly arose from a local stimulation of transglutaminase synthesis and secretion in fibroblasts, pointing out that the differences he was able to detect in the acute (children) and chronic (adult) disease might be related to the prevalence of apoptotic events in long-lasting CD. In this respect he also interestingly pointed out the possibility of different glutamine deamidation mechanisms, involving bacterial rather than tissular Tgase, to generate the deamidated gliadin epitopes recognized by intestinal T-cells. Ilma Korpanay-Szabo (University of Tampere and Children Hospital, Budapest) further described studies on the immunoreactivity of intestinal mucosa, employing immunocytochemical staining with a panel of antiTGase monoclonal antibodies (as well as IgA sera derived from CD patients) proving that the monoclonal antibodies have different staining patterns, which might eventually depend on the exposition of selective conformational epitopes, triggered by local interaction with ligands and/or other macromolecules. Briefly, the data indicate that fibronectin is the anchor for antiTGase immunoreactivity and that CD patients tested antibodies recognized the enzyme in a conformation dependent way, with high reactivity for Ca++ activated transglutaminase, deposing massively in the flat intestinal mucosa. Additional ligands (Mg, GTP, GDP and calreticulin) have less evident effects on tissular Tgase immunoreactivity. In his presentation Roberto Marzari (University of Trieste) described very careful studies to identify the regions of tissular transglutaminase reactive against IgA autoantibodies produced by intestinal lymphocytes, as determined by a phage display library and a single chain antibody fragment approach. The approach they used employed truncation mutants of recombinant human transglutaminase lacking controlled regions at the N-and the C-terminal end of the peptide chain, even under conditions of intracellular expression, in a characteristic reducing milieu, which usually interferes with correct antibody folding, by the IACT (Intracellular Antibody Capture Technology), quite promizing for gene therapy and functional genomics. The presentation by Roberto Troncone (University of Naples) focused on the effects of transglutaminase-directed antibodies on the activity of tTGase both in situ and in vitro, demonstrating a moderate, but significant, inhibition of crosslinking activity by CD patientsí antisera, without clear correlation with the antibody titre detected by standard ELISA assays,as it could be predicted on the basis of a heterogenity in the recognized epitopes. They further reported effects of calcium which differed significantly from those described in additional communication, further enlighting the possibility of heteregeneous antibody populations. Tuula Halttunen (University of Tampere) further added onto the biological effects of IgA antibody fractions, isolated from CD patients, in inhibiting intestinal wall differentiation in an vitro model of cultured T84 cells. The tested antibodies interfered with cell aggregation processes, which are believed to mimick in vitro the ontogenesis of the intestinal crypts. The same effects were observed with monoclonal antibodies (as CUB 7402) directed against tissular TGase, suggesting that interference with TGase related processes, e.g. BM processing, might interfer with normal intestinal ontogenesis or regeneration in CD patients.
The final session of the workshop was centered on the mechanism of TGase involvement in disruption of tissular homeostasis, largely through its involvement in programmed cell death, since these effects might be relevant in the pathogenesis of CD. Thus Wilbert Boelens (University of Nijmegen) examined cellular accumulation of two rather ubiquitous chaperonin proteins (aB-crystallin and HSP-27) which are also substrates of TGase, as both amino and acyl donor. This is particularly true for HSP27 which can interact with cytoskeletal proteins, particularly actin and vimentin, forming complexes in which both contributing proteins become substrates of transglutaminase with crosslinkage through isopeptide bonds. It is believed that this is the mechanism by which pathologic inclusion bodies develop in several neurologic diseases as well as in alcoholic hepatopathy. In this perspective Laszlo Fesus (University of Debrecen) reviewed research activity, directed to elucidate the role of tTGase in programmed cell death. The TGase gene expression is definitely induced in apoptotic cells, through the interaction of apoptotic signalling molecules with unidentified regions in the promoter and TGase becomes active in the intracellular compartment, depending on Ca++ and GTP levels, as verified by steady-state levels of circulating isopeptide bonds. Apparently cells undergo apoptosis, following triggering by external signals, by different pathways which require massive stimulation of caspases or of TGases, and removal of apoptotic cells by phagocytosis, without release of intracellular materials including TGase, to avoid surges of autoantibodies. This is generally consistently dealt with, but in CD. Data by Dr. Fesus on effects of antibodies on tTGase activity (lack of effects, activation or inhibition in different serum samples) were consistent with a great heterogeneity in the antibody response, as evident also in the words of other speakers. The final lecture of the meeting was given by Mauro Piacentini (University of Rome-Tor Vergata) who provided evidence on the distribution of transglutaminase activity and protein immunoreactivity within the intestinal wall of control and CD patients, the occurrence of typical signs of apoptotic cells within the intestinal mucosa, but also evidences that similar hystologic and immunocytochemical changes can be observed also in other inflammatory intestinal diseases, bringing about a word of caution in the interpretation of ongoing research.