Abstract
Based on previous results demonstrating that complexes of a mutant α1-antitrypsin with the heat shock proteins (HSP)70 and glucose-regulated protein94 (Grp94) circulate in the blood of patients with type 1 diabetes, we raised the hypothesis that these complexes could represent the primary antigen capable of triggering the autoimmune reactions leading to overt diabetes. As a first approach to this issue, we searched whether A1AT and HSPs had a sequence similarity to major islet antigen proteins so as to identify among the similar sequences those with potential relevance for the pathogenesis of diabetes. A thorough in silico analysis was performed to establish the score of similarity of the human proteins: A1AT, pro-insulin (INS), GAD65, IAPP, IA-2, ICA69, Grp94, HSP70 and HSP60. The sequences of A1AT and HSPs with the highest score of similarity to the islet peptides reported in the literature as the main autoantigens in human diabetes were recorded. At variance with other HSPs, also including HSP90 and Grp78, Grp94 contained the highest number and the longest sequences with structural similarity to A1AT and to well-known immunogenic peptides/epitopes of INS, GAD65, and IA-2. The similarity of A1AT with Grp94 and that of Grp94 with INS also suggested a functional relationship among the proteins. Specific sequences were identified in A1AT, Grp94 and HSP70, with the highest score of cross-similarity to a pattern of eight different islet protein epitopes. The similarity also involved recently discovered autoantigens in type 1 diabetes such as a hybrid peptides of insulin and the defective ribosomal insulin gene product. The significant similarity displayed by specific sequences of Grp94 and A1AT to the islet peptides considered main antigens in human diabetes, is a strong indication for testing these sequences as new peptides of immunogenic relevance in diabetes.
Similar content being viewed by others
Introduction
Despite strenuous efforts to define the aetiology and pathogenic mechanism of type 1 diabetes (T1D), the precise cause of the disease remains unknown, and no specific cure is available1. None of the many factors identified thus far as causally related to diabetes, including the major islet cell auto-antigens, can explain the heterogeneous molecular and clinical features of the disease2. Thus, although antigen-specific immune-therapies have not met expectations3,4,5, the recognition that a wide range of different mechanisms concurs with beta cells destruction has prompted investigators to focus on broad-spectrum, non-antigen-specific, immune-based therapies6. Special focus has been placed on α1–antitrypsin (A1AT), the main circulating anti-proteinase inhibitor, which is used as a diabetes therapeutic agent for its broad anti-inflammatory and immune-modulatory properties unrelated to the specific proteinase inhibitory activity7. Although clinical trials of A1AT have not been successful8, the finding that A1AT is expressed in both islet9 and endothelial cells10 of the human pancreas, in addition to its protective effects on beta cells11,12 and islet transplantation13, corroborated the idea that A1AT actually plays a role in regulating islet cell homeostasis and glucose metabolism, confirmed by studies on genes related to the metabolic syndrome14.
The A1AT gene undergoes a high mutation rate: thus, there are many variants of the protein, with some also related to specific pathologies15. When mutations severely affect A1AT folding, so as to prevent protein secretion, the endoplasmic reticulum (ER) machinery employing specific heat shock proteins (HSPs) is set up to restore protein folding16. If this process fails, the mutated A1AT (mA1AT) is directed to the ER-associated-degradation pathway with the primary assistance of glucose-regulated protein94 (Grp94)17, the most represented ER HSP. The resulting intense ER stress that follows activation of the unfolded protein response (UPR) in A1AT-expressing cells is responsible for cell apoptosis16 and extracellular liberation of the protein content.
In previous works, we demonstrated that a variant, probably a mutated form of A1AT that is structurally and functionally distinct from normal A1AT (wA1AT), was present in the plasma of patients with T1D18 and linked to HSP70 in complexes with Grp9419,20. This finding was the unequivocal sign of an upstream defective folding mechanism of mA1AT eventually responsible for the intracellular events that result in the engagement and extracellular liberation of HSPs. Considering also that HSPs outside the cell are a potent immunogenic stimulus, and that HSPs are found in the plasma of diabetic subjects even years after diabetes onset20, we hypothesised that complexes of mA1AT with HSPs could represent a true, stable antigen in diabetes, that is responsible for initiating and maintaining autoimmune reactions. As a first investigative approach to this problem, we sought to determine whether A1AT and HSPs share a structural similarity to the pancreas islet proteins already identified as the main antigens in diabetes, and, in particular, whether this similarity indicates that the islet peptides are key players in autoreactivity and loss of tolerance in human diabetes. The comprehensive and thorough in silico analysis subsequently conducted on sequences of A1AT, HSPs and the main islet proteins provides preliminary information about the potential immunogenic implication of A1AT/HSP complexes in the pathogenesis of T1D.
Methods
Human proteins were recovered through UniProtKB/Swiss-Prot (http://www.uniprot.org). Proteins (entry name with the accession number) were: alpha1 antitrypsin (A1AT, P01009-1); Pro-Insulin (INS, P01308-1); Islet Cell Autoantigen 1 (ICA69, Q05084); Glutamate Decarboxylase 2, GAD65 (DCE2, Q05329); Receptor-type Tyrosine-protein Phosphatase-like N, IA-2 (PTPRN, Q16849); Islet Amyloid Polypeptide (IAPP, P10997); Endoplasmin, Grp94 (ENPL, P14625); Heat Shock 70 kDa protein 1A (HS71A, P0DMV8); and 60 kDa HSP (CH60, P10809). Also retrieved were the sequences of human: Endoplasmic reticulum chaperone Bip (Grp78; P11021), Heat Shock Protein 90-alpha (HSP90AA1-1, P07900-1), Heat Shock Protein 40 (DNAJB1, P25685), Zinc transporter 8 (ZnT8, Q8IWV4), Glucose 6-phosphatase 2 (IGRP, Q9NQR9), Urocortin 3 (UCN3, Q969E3), Insulin gene enhancer protein isl-1 (ISL-1, P61371).
To detect sequence similarities, pairwise alignment of the proteins was performed using two progressive alignment methods, the Clustal Omega programme (UniProtKB) and the T-Coffee (Tree-based Consistency Objective Function for alignment Evaluation) package that combines different sequence-alignment programmes21. The two methods use different algorithms for sequence alignment; thus, different sequences can be detected as similar in the same protein. Although T-Coffee has proven to be superior to other methods, Clustal Omega was combined in any pairwise sequence alignment to detect as many similar sequences as possible, a condition not satisfied by a single method. The concordance of both methods in detecting sequence similarity added further relevance to the sequence similarity. The programmes chosen in the T-Coffee package to obtain consensus alignment were muscle msa (multiple sequence alignment) and T-Coffee msa, both of which start by computing pairwise alignments of the input sequences. The parameter setting included regular computation mode. Only sequences of at least eight consecutive amino acid residues without any gap and with the highest consistency score were considered in the computation.
In the first step (Fig. 1), A1AT was aligned with any of the islet proteins considered a main autoantigen in T1D, namely, pro-insulin (INS), GAD65, IA-2, ICA69, IAPP and HSP60, and with Grp94 and HSP70, the two HSPs found in complexes with mA1AT in the plasma of diabetic patients19,20. HSP60 was included in the analysis since, similarly to other main islet protein antigens, it represented a source of immunogenic peptides, also used as vaccine in human diabetes22. The sequences of A1AT with strong similarity to both islet proteins and HSPs were analysed using the msa programme of T-Coffee to test the overall score of cross-similarity. In parallel, Grp94, HSP70 and HSP60 were aligned with INS, GAD65, IA-2, ICA69 and IAPP to detect similar sequences (Fig. 1). The analysis was then centred on the sequences of both A1AT and HSPs with similarity to all the peptides/epitopes of INS, GAD65, ICA69, IA-2 and IAPP reported in the literature as immunogenic in human diabetes. By taking into consideration the sequences of HSPs similar to both islet peptides/epitopes and A1AT, it was possible to identify other sequences in A1AT with cross-similarity to islet epitopes besides those recorded in pairwise alignment with any islet protein (Fig. 1). All of the sequences of A1AT similar to both HSPs and islet protein epitopes were listed, and those with the highest similarity score (100%) to two or more islet protein epitopes were analysed using T-Coffee (https://www.ebi.ac.uk/Tools/msa/tcoffee) and Clustal Omega (https://www.ebi.ac.uk/Tools/msa/clustalo/) which also provide information about the evolutionary relationship of the protein sequence.
To further prove the specificity of sequence similarity, pairwise and msa with the islet proteins and HSPs were also performed with the sequences of human serum albumin (HSA, P02768), bovine serum albumin (BSA, P02769) and vitamin D binding protein (VTDB, P02774) taken as proteins of reference.
Results
Sequence similarity between A1AT and islet proteins encompasses similarity to HSPs
Among the proteins considered in the alignment with A1AT (Fig. 2), Grp94 showed the most extensive similarity (Tables S1 and S2), whereas fewer similar sequences were found in both HSP70, in which similar sequences were mostly found in the first 440 residues, and HSP60, in which similarity regarded the sequences 398–413 and 431–449 of HSP60 (Fig. 2; Tables S1 and S2) identified as immunogenic peptides22.
The islet proteins with extensive similarity to A1AT were INS, GAD65 and IA-2 (Fig. 2; Tables S1 and S2). Almost the entire INS sequence (105 of 110 residues), including all of the known epitopes of INS23,24,25,26,27,28,29,30,31,32,33 (Fig. 3A), showed similarity to A1AT (Table 1). In particular, the INS sequences 18–39, 48–70 and 72–107 were similar to distinct A1AT sequences, with the longer ones being A1AT 270–291 and 300–325 (Fig. 2, 3A). Unlike INS, IAPP did not show any extensive sequence similarity to A1AT (Fig. 2; Table S1), although three separate A1AT sequences showed similarity to the immunogenic peptides of IAPP29,34 (Fig. 3B).
Of relevance was the broad similarity shared by A1AT with GAD65, because it comprised almost half of the GAD65 residues (Table S2) with very long sequences (Fig. 2), some of which covered the majority of the known GAD65 epitopes23,26,27,29,35, including those in the 473–543 immune-dominant peptide35 (Fig. S1, Fig. 4A). Importantly, most of the GAD65 sequences similar to A1AT were detected by both alignment methods (Fig. 2).
In addition, in the case of IA-2, there was agreement among the methods in identifying the sequences of IA-2 similar to A1AT as those prevalently located in the cytosolic domain (601–979). Only T-Coffee found four additional similar sequences in the luminal portion of IA-2 (35–575) and another single sequence in the trans-membrane fragment (576–600) (Fig. 2; Table S1). As a result, very long portions of IA-2, mainly in the luminal and trans-membrane domains, were not aligned at all with A1AT, with 276 of 979 residues being thus similar to A1AT (Table S2).
Not so significant was the similarity shared by A1AT with ICA69 (Fig. 2; Table S2) although in this case, the similarity also comprised the epitopes of ICA69 predicted to bind to HLA molecules, (i.e., sequences 69–78 and 222–231; Table S1)26.
The overview of the alignments of any protein with A1AT (Fig. 2) demonstrated that many sequences of islet proteins shared with HSPs a similarity to the same sequences of A1AT. Thus, to test in msa the extent of cross-similarity, we grouped the sequences of A1AT similar to at least two other proteins of approximately the same length (Fig. S1). Most A1AT sequences (17 of 25) showed significant cross-similarity to Grp94, either alone or with HSP70 or HSP60, and, in decreasing order of frequency, to GAD65, IA-2, INS, ICA69 and IAPP (Fig. S1). Of interest was the long 144–168 sequence of A1AT with similarity to GAD65 182–206, IA-2 699–716 and Grp94 205–229 (Fig. S1). On the whole, the sequences of A1AT with similarity to the higher number of islet proteins and HSPs were around amino acid (a.a.) 300–320 and 380–400 (Fig. 2, Fig. S1).
Similarity of HSPs to islet proteins supports broadening of cross-reactivity to A1AT
The finding that many sequences of the islet proteins shared with HSPs, especially Grp94, similarity to A1AT (Fig. 2; Fig. S1) allowed us to hypothesise a broader similarity between HSPs and islet proteins. We investigated this issue by aligning each HSP with islet proteins and recording, among the numerous similar sequences of HSPs (not shown), those with similarity to the immunogenic peptides/epitopes of any islet protein (Fig. 1, steps on right).
As observed with A1AT, also in the alignment with HSPs the sequence of INS was almost entirely covered by similar sequences of Grp94, HSP70 and HSP60 (Table 1). Excluding the 17–24 epitope in the signal peptide, similar to a single sequence of HSP70 and HSP60 (Fig. 3A), the remaining INS sequence, comprising C-peptide and A-chain, showed similarity especially to Grp94 and HSP70. Of note was the long 611–656 sequence of Grp94 with similarity to the 66–107 sequence of INS, identified by both the alignment methods as two separate, partly overlapping peptides (Fig. 3A).
In the pairwise alignment with IAPP, both Grp94 and HSP60 contained sequences similar to almost the entire IAPP sequence (data not shown), including the known epitopes29,34 (Fig. 3B), whereas only two single sequences of HSP70 were found similar to IAPP in the portion comprising the 65–84 immunogenic peptide29 (Fig. 3B).
In the alignment of HSPs with GAD65, Grp94 showed similarity to all known GAD65 epitopes, with the only exception of the peptide 552–57023 similar to HSP70 (Fig. 4A). Of particular relevance was the long 683–740 Grp94 sequence similar to GAD65 498–555, because part of this sequence that comprised two GAD65 epitopes, also showed similarity to HSP70 573–603 and to the 443–461 immunogenic peptide of HSP6022 (Fig. 4A). The sequences of Grp94, HSP70 and HSP60 similar to GAD65 epitopes also displayed cross-similarity to the sequence 384–398 of A1AT already found similar to the same GAD65 peptide 509–523 in the alignment of A1AT with GAD65 (Fig. 2, Fig. S1) and to IAPP sequence 74–88 (Fig. S1). The sequences of A1AT similar to those of HSPs with similarity to islet epitopes turned out to display cross-similarity to the same islet epitopes when analyzed in msa (Fig. 4, asterisks on A1AT).
Also in the alignment with IA-2 most of the numerous similar sequences found in HSPs belonged to Grp94, whose sequence covered six of the nine IA-2 epitopes23,26,29,36 (Fig. 4B). Thus, the structural similarity to the epitopes of IA-2 was apparently better sustained by Grp94 and HSP70 than by A1AT (Table S1).
Both alignment methods detected an extensive sequence similarity of each HSP with ICA69 (data not shown) including all the sequences identified as immunogenic (Fig. 4C)26. This result differed from that obtained with A1AT, which showed limited similarity to ICA69 epitopes (Table S1). However, cross-similarity to ICA69 epitopes was displayed by the A1AT sequences highly similar to HSPs (Fig. 4C, asterisks on A1AT).
To exclude that the broad structural similarity displayed by Grp94 and HSP70 with islet protein epitopes was not specific, i.e., it could also be shared by other HSPs, we extended the investigation to the HSPs with extensive structural similarity to Grp94 and HSP70. Neither HSP90 (80% sequence similarity with Grp94) nor Grp78 (90% sequence similarity with HSP70) showed any sequences with similarity to the epitopes of GAD65 and ICA69, whereas in the pair-wise alignment with IA-2, only a single sequence in HSP90 (505–518), but none in Grp78, was found similar to a single epitope of IA-2 (792–805). An even more impressive difference, with respect to what observed with Grp94 and HSP70, was noted in the alignment of HSP90 and Grp78 with INS, since only two short and distant sequences in INS were covered by similar sequences of both HSP90 and Grp78 (data not shown).
Sequences of A1AT and HSPs with the best score of similarity to different islet epitopes
To identify the sequences of both A1AT and HSPs that might better satisfy the requisites of potential epitopes in diabetes, we considered only those with similarity to islet protein epitopes (Table 2). Twenty-six sequences of A1AT together with a higher number of Grp94 sequences, among those of HSPs, shared mutual similarity to the indicated epitopes of IA-2, INS, GAD65, IAPP and ICA69 (Table 2). As expected from the cross-alignments of both A1AT and HSPs with islet proteins, and of A1AT with HSPs (Figs. 2, 3, 4), the same A1AT sequence showed similarity to distinct islet epitopes and to distinct HSP sequences (Table 2). We chose the sequences of A1AT that, in addition to adequately matching the length of the corresponding islet epitopes, also showed cross-similarity to at least two HSPs and two islet epitopes. The sequences were: A1AT 212–224, 305–321, 363–376 and 384–398 (Table 2). These groups of sequences were analysed in depth after the best alignment with the perfect score of similarity was guaranteed in each group. To this aim, the sequences of HSP70 293–304, ICA69 283–293 and HSP60 398–409 were omitted from the analysis, as they did not support the required extent of similarity and alignment in the group to which they belonged (Fig. 5A). In msa, the four A1AT sequences, considered both alone and together with the corresponding similar sequences of Grp94 and HSP70, were perfectly similar to each other (Fig. 5b, left), implying that each group of sequences could be equally similar to all islet epitopes. This was also confirmed by the unexpected finding that all islet epitopes were highly similar to each other (Fig. 5B, right), suggesting a common structural denominator of both these islet epitopes and the sequences of A1AT, Grp94 and HSP70.
Looking deeper into how these A1AT and HSP sequences could be related to each other in supporting similarity to the islet epitopes, any single group, as well as all the four groups of A1AT and HSP sequences, were analysed together with all islet epitopes in msa, and the score of similarity and evolutionary relationship were recorded (Fig. 5B). The groups of A1AT and HSP sequences considered together were perfectly aligned to all islet epitopes with the maximal similarity score reached after excluding the IA-2 epitope. This epitope was perfectly aligned with A1AT 212–224, 364–376 and 384–398, but not equally well with A1AT 308–320 (Fig. 5B). The evolutionary analysis of the sequences indicated that A1AT 308–320 and Grp94 639–651 were, respectively, closely associated with INS 90–102 and 59–71 and had a recent common ancestor (data not shown). It is worth noting that none of the reference proteins was able to support the similarity to all islet epitopes as did the above sequences of A1AT and HSPs.
To ascertain whether other islet proteins related to pathogenesis of human diabetes could support structural similarity to A1AT and HSPs as did the main islet protein epitopes, we considered the epitopes identified in ZnT826,37, IGRP38, UCN337, ISL-137 as well as the immunogenic peptide (DRiP) derived from an alternative open reading frame within the human insulin mRNA with the capacity to sustain autoimmunity in clinical diabetes39. Whereas no peptide, either epitope or not, of the above islet proteins showed similarity to any of the above sequences of A1AT, Grp94 and HSP70 (Figs. 4, 5A), a high score of cross-similarity to some of these A1AT and HSP sequences was instead displayed by DRiP due to its main similarity to INS 1–10 (Fig. S2). Since DRiP was also found responsible for killing beta cells39, the structural similarity to the INS-like epitope appears of particular relevance for driving antigenicity in human diabetes.
Sequences of A1AT and HSPs that better satisfy the requisite of antigenic peptides in T1D
In consideration of the above results, we extended the investigation also to the recently identified neo-antigens, hybrid insulin peptides (HIPs), that are products of the post-translational modification of islet proteins40,41. HIPs are formed with part of the C-peptide fused with either pancreatic neuropeptide Y (NNPY), IAPP or A-chain and are key antigens for auto-reactive T cells in T1D29,42. We tested whether the HIPs that activate human CD4 T cells showed similarity to the sequences of both A1AT and HSPs with cross-similarity to islet epitopes (Fig. 5B). We noted that HIPs, regardless of whether their right part was from NNPY, IAPP or A-chain, were perfectly similar to each other and that the sequences of A1AT, Grp94 and HSP70 similar to all the islet peptides (Fig. 5B) had the first seven residues similar to the peptides of the right part in HIPs and the last eight residues similar to the common (GQVELGGG) peptide of each HIP, the two peptides of each A1AT and HSP sequence having the two central residues in common (Fig. 5C). We thus tested in pairwise alignment and in msa the extent of similarity between each HIP and the peptides obtained by fusing the sequences of A1AT with those of either Grp94 or HSP70, with each sequence being similar to either the right or left part of any HIP, so as to satisfy all the possible combinations in each group of A1AT and HSP sequences (Fig. 5C). The A1AT/HSP peptides that best aligned with HIPs with the highest score of similarity also in msa were from A1AT 384–396 with Grp94 694–704 and from A1AT 308–320 with Grp94 639–651 (Fig. 5C). This Grp94 sequence was also evolutionarily closely related to INS 90–102 (data not shown). Only these last A1AT and Grp94 sequences contained the peptides that in different combinations were perfectly similar to all HIPs (Fig. 5C).
Discussion
Based on previous results indicating that complexes of mA1AT with HSP70 and Grp94 found in the plasma of type 1 diabetic subjects20,43 might be markers of disease with potential immunogenic relevance for disease development, we explored here the possibility that A1AT and HSPs display structural similarity to the peptides already identified in islet proteins as immunogenic in human diabetes. If the assumption is true that mA1AT is the very first trigger of intracellular events causally related to HSP association and extracellular liberation16,17,44 (Fig. 6), identifying the peptides that in the complex have strong structural similarity to already known antigens in diabetes can help to prove the immunogenic potential of these new peptides. Thus, although the in silico analysis is preliminary and partial, it might be useful for providing the necessary indications.
Our results first demonstrate a significant, broad sequence similarity between A1AT and Grp94 (Fig. 2; Table S1), which likely predicts the spreading to A1AT of the reactivity arising against the HSP. In fact, HSPs outside the cell are highly immunogenic, and in particular Grp94 are found in the plasma of diabetic patients, even years after diabetes onset20. Thus, HSPs could be primarily responsible for any autoimmune reaction triggered by the complex, also demonstrated by the fact that anti-HSP Abs circulate in the plasma of diabetic patients42. Cross-reactivity of anti-HSP Abs with many structurally similar portions of A1AT can also account for the reduced plasma concentration and defective functioning of A1AT frequently observed in T1D45,46.
Implication of the mA1AT/HSP complex in the autoimmune process related to diabetes is supported by the significant structural similarity shared by both HSPs, Grp94 in particular, and A1AT with all the peptides/epitopes already identified in the major islet antigen proteins (i.e., INS, GAD65 and IA-2; Figs. 3, 4), a condition not satisfied instead by any of the reference proteins nor by any of the HSPs that show a high degree of structural similarity with Grp94 and HSP70. Particularly striking was the result of almost complete coverage of the INS sequence by distinct similar sequences of A1AT and by very long sequences of Grp94 (Table 1; Fig. 3A). The possibility that the structural similarity shared by A1AT and Grp94 with INS also implies a functional link is supported by the fact that the A1AT expression is regulated by hepatocyte nuclear factors (HNFs) which are involved in the regulation of many genes in human islets47, and that some HNF variants are related to the development of monogenic forms of diabetes48. Further evidence also comes from the functional link that exists between the A1AT expression and a variant form of HNF1A associated with impaired insulin secretion14 and by the demonstration that Grp94 has an essential role in regulating beta cell function and pro-insulin level in human beta cells49.
In line with these considerations are the results of the alignments of A1AT and HSPs with GAD65 and IA-2 (Fig. 4). More than 60% of A1AT residues displayed similarity with both GAD65 and IA-2 (Table S2) in long sequences (≥ 15 residues), whereas, among HSPs, Grp94 contained a higher number of sequences and residues (450) similar to GAD65. Of particular interest was the Grp94 sequence of 58 residues covering two separate epitopes of GAD65, also similar to two separate sequences of A1AT (Fig. 4A). In the case of IA-2, similarities with A1AT were found almost always exclusively in the cytoplasmic segment of IA-2 which undergoes cleavage and translocation to the nucleus where it enhances pancreatic beta cell proliferation and promotes the expression of insulin genes50. The similarity of A1AT with the cytoplasmic IA-2 fragment further suggests A1AT’s functional involvement in the pathway that links insulin secretion with gene expression in beta cells.
Similar considerations can be applied to the extensive sequence similarity found between IA-2 and HSPs, characterised by the remarkable length of some sequences (> 40 residues) distributed rather homogeneously along the entire IA-2 sequence. Compared with HSP70 and HSP60, Grp94 again displayed a greater number of similar residues (486) prevalently covering the trans-membrane and cytoplasmic fragments of IA-2 with very long sequences. A recent study proposed that autoimmunity triggered by GAD65 and IA-2 is linked to the release of these proteins in exosomes in association with ER chaperones, including Grp9451. The conclusion that immunogenicity is primarily linked to the HSPs leaving the ER under stress conditions corroborates our results that indicate that a broad structural similarity can justify the spreading of autoimmunity from HSPs to other islet proteins (Fig. 6).
Because many sequences of A1AT and HSPs showed similarity to the major islet antigens (Fig. S1; Table 2), we identified the sequences that might best constitute immunogenic peptides. Among all A1AT sequences partly overlapping each other in covering separate islet epitopes, we considered only those with similarity to at least two islet epitopes and HSPs; intriguingly, these A1AT sequences were highly similar to each other and the corresponding Grp94 and HSP70 sequences (Fig. 5B). Since also the epitopes of INS, GAD65, IA-2, IAPP and HSP60 were highly similar to each other (Fig. 5B) it followed that the A1AT and HSP sequences, taken both separately and together, shared similarity with all the islet epitopes (Fig. 5B), a finding suggesting that a common structural denominator was shared by A1AT and HSPs with these islet epitopes. We verified whether this conclusion could be extended to other autoantigens of islet proteins, as well as to newly discovered peptides of insulin targeted by reactive T-cells in human diabetes, such as the hybrid insulin products (HIPs)29,42 and the defective immunogenic peptide translated from the insulin mRNA39. Negative results were obtained by testing the epitopes of ZnT8, IGRP, UCN3 and ISL-1 whereas the insulin-related epitopes, both non-conventional (Fig. S2) and hybrid (Fig. 5C) turned out to display strong similarity to the A1AT/HSP sequences with similarity to the panel of the main islet peptides/epitopes (Fig. 5B). Apparently thus, the structural similarity to specific INS epitopes was crucial for supporting cross-similarity of the A1AT and HSP sequences to other main islet antigens of GAD65 and IA-2. Although sequence similarity does not equate to functional similarity, by investigating protein–protein interaction in String database, we observed that a close association, implying functional relationship, occurred between A1AT and Grp94, but not other HSPs (Fig. S3) and that INS covered a central node that functionally linked two separate clusters of proteins, one comprising A1AT together with the HSPs, the other formed with GAD65 and IA-2 characterized by a closer functional linkage to INS (Fig. S3).
Data availability
All data are available from the authors on request.
Abbreviations
- A1AT:
-
Alpha 1-antitrypsin
- a.a.:
-
Amino acid
- BSA:
-
Bovine serum albumin
- ER:
-
Endoplasmic reticulum
- HIP:
-
Hybrid insulin peptides
- HSA:
-
Human serum albumin
- HSP:
-
Heat Shock Protein
- msa:
-
Multiple sequence alignments
- UPR:
-
Unfolded protein response
- VTDB:
-
Vitamin D binding protein
References
von Herrath, M. G., Korsgreen, O. & Atkinson, M. A. Factors impeding the discovery of an intervention-based treatment for type 1 diabetes. Clin. Exp. Immunol. 183, 1–7 (2015).
Norris, J. M., Johnson, R. K. & Stene, L. C. Type 1 diabetes-early life origins and changing epidemiology. Lancet Diabetes Endocrinol. 8(3), 226–238 (2020).
Roep, B. O., Wheeler, D. C. S. & Peakman, M. Antigen-based immune modulation therapy for type 1 diabetes: The era of precision medicine. Lancet Diabetes Endocrinol. 7(1), 65–74 (2019).
Michels, A. W. & Gottlieb, P. A. Learning from past failures of oral insulin trials. Diabetes 67(7), 1211–1215 (2018).
Kolb, H. & von Herrath, M. Immunotherapy for type 1 diabetes: Why do current protocols not halt the underlying disease process?. Cell Metab. 25(2), 233–241 (2017).
Atkinson, M. A., Roep, B. O., Posgai, A., Wheeler, D. C. S. & Peakman, M. The challenge of modulating β-cell autoimmunity in type 1 diabetes. Lancet Diabetes Endocrinol. 7(1), 52–64 (2019).
Jonigk, D. et al. Anti-inflammatory and immunomodulatory properties of α1-antitrypsin without inhibition of elastase. Proc. Natl. Acad. Sci. USA 110(37), 15007–15012 (2013).
Lebenthal, Y. et al. A phase II, double-blind, randomized, placebo-controlled, multicenter study evaluating the efficacy and safety of alpha-1 antitrypsin (AAT) (glassia) in the treatment of recent-onset type 1 diabetes. Int. J. Mol. Sci. 20, 6032 (2019).
Bosco, D. et al. Expression and secretion of alpha1-proteinase inhibitor are regulated by proinflammatory cytokines in human pancreatic islet cells. Diabetologia 48(8), 1523–1533 (2005).
Lou, J. et al. Expression of alpha-1 proteinase inhibitor in human islet microvascular endothelial cells. Diabetes 48(9), 1773–1778 (1999).
Koulmanda, M. et al. Alpha 1-antitrypsin reduces inflammation and enhances mouse pancreatic islet transplant survival. Proc. Natl. Acad. Sci. USA 109(38), 15443–15448 (2012).
Guttman, O. et al. α1-Antitrypsin modifies general NK cell interactions with dendritic cells and specific interactions with islet β-cells in favor of protection from autoimmune diabetes. Immunology 144, 530–539 (2014).
Baranovski, B. M. et al. Exploration of α1-antitrypsin treatment protocol for islet transplantation: Dosing plan and route of administration. J. Pharmacol. Exp. Ther. 359(3), 482–490 (2016).
Setoh, K. et al. Three missense variants of metabolic syndrome-related genes are associated with alpha-1 antitrypsin levels. Nat. Commun. 6, 7754 (2015).
Chappell, S. et al. A polymorphism of the alpha1-antitrypsin gene represents a risk factor for liver disease. Hepatology 47(1), 127–132 (2008).
Guerriero, C. J. & Brodsky, J. L. The delicate balance between secreted protein folding and endoplasmic reticulum-associated degradation in human physiology. Physiol. Rev. 92(2), 537–576 (2012).
Christianson, J. C., Shaler, T. A., Tyler, R. E. & Kopito, R. OS-9 and GRP94 deliver mutant alpha1-antitrypsin to the Hrd1-SEL1L ubiquitin ligase complex for ERAD. Nat. Cell Biol. 10(3), 272–282 (2008).
Finotti, P., Carraro, P. & Calderan, A. Purification of proteinase-like and Na+/K(+)-ATPase stimulating substance from plasma of insulin-dependent diabetics and its identification as alpha 1-antitrypsin. Biochim. Biophys. Acta 1139(1–2), 122–132 (1992).
Finotti, P. & Pagetta, A. A heat shock protein70 fusion protein with alpha1-antitrypsin in plasma of type 1 diabetic subjects. Biochem. Biophys. Res. Commun. 315(2), 297–305 (2004).
Pagetta, A., Folda, A., Brunati, A. M. & Finotti, P. Identification and purification from the plasma of type 1 diabetic subjects of a proteolytically active Grp94. Evidence that Grp94 is entirely responsible for plasma proteolytic activity. Diabetologia 46(7), 996–1006 (2003).
Notredame, C., Higgins, D. G. & Heringa, J. T-Coffee: A novel method for fast and accurate multiple sequence alignment. J. Mol. Biol. 302(1), 205–217 (2000).
Horváth, L. et al. Antibodies against different epitopes of heat-shock protein 60 in children with type 1 diabetes mellitus. Immunol. Lett. 80(3), 155–162 (2002).
van Lummel, M. et al. Dendritic cells guide islet autoimmunity through a restricted and uniquely processed peptidome presented by high-risk HLA-DR. J. Immunol. 196(8), 3253–3263 (2016).
Pathiraja, V. et al. Proinsulin-specific, HLA-DQ8, and HLA-DQ8-transdimer-restricted CD4+ T cells infiltrate islets in type 1 diabetes. Diabetes 64(1), 172–182 (2015).
Michels, A. W. et al. Islet-derived CD4 T cells targeting proinsulin in human autoimmune diabetes. Diabetes 66(3), 722–734 (2017).
van Lummel, M. et al. Posttranslational modification of HLA-DQ binding islet autoantigens in type 1 diabetes. Diabetes 63(1), 237–247 (2014).
Rudy, G. et al. Similar peptides from two beta cell autoantigens, proinsulin and glutamic acid decarboxylase, stimulate T cells of individuals at risk for insulin-dependent diabetes. Mol. Med. 1(6), 625–633 (1995).
Wan, X. & Unanue, E. R. Unique features in the presentation of insulin epitopes in autoimmune diabetes: An update. Curr. Opin. Immunol. 46, 30–37 (2017).
Babon, J. A. B. et al. Analysis of self-antigen specificity of islet-infiltrating T cells from human donors with type 1 diabetes. Nat. Med. 22(12), 1482–1487 (2016).
Kent, S. C. et al. Expanded T cells from pancreatic lymph nodes of type 1 diabetic subjects recognize an insulin epitope. Nature 435(7039), 224–228 (2005).
Wan, X., Thomas, J. W. & Unanue, E. R. Class-switched anti-insulin antibodies originate from unconventional antigen presentation in multiple lymphoid sites. J. Exp. Med. 213(6), 967–978 (2016).
Savinov, A. Y., Wong, F. S., Stonebraker, A. C. & Chervonsky, A. V. Presentation of antigen by endothelial cells and chemoattraction are required for homing of insulin-specific CD8+ T cells. J. Exp. Med. 197(5), 643–656 (2003).
Alhadj Ali, M. et al. Metabolic and immune effects of immunotherapy with proinsulin peptide in human new-onset type 1 diabetes. Sci Transl Med 9, eaaf7779 (2017).
Panagiotopoulos, C., Qin, H., Tan, R. & Verchere, C. B. Identification of a beta-cell-specific HLA class I restricted epitope in type 1 diabetes. Diabetes 52(11), 2647–2651 (2003).
Lohman, T. et al. Immunodominant epitopes of glutamic acid decrboxylase 65 and 67 in insulin-dependent diabetes mellitus. Lancet 343(8913), 1607–1608 (1994).
Bearzatto, M. et al. Two distinctly HLA-associated contiguous linear epitopes uniquely expressed within the islet antigen 2 molecule are major autoantibody epitopes of the diabetes-specific tyrosine phosphatase-like protein autoantigens. J. Immunol. 168(8), 4202–4208 (2002).
Gonzalez-Duque, S. et al. Conventional and neo-antigenic peptides presented by β cells are targeted by circulating naïve CD8+ T cells in Type 1 diabetic and healthy donors. Cell Metab. 28(6), 946–960 (2018).
Mallone, R. et al. CD8+ T-cell responses identify β-cell autoimmunity in human type 1 diabetes. Diabetes 56, 613–621 (2007).
Kracht, M. J. L. et al. Autoimmunity against a defective ribosomal insulin gene product in type 1 diabetes. Nat. Med. 23(4), 501–507 (2017).
Kolb, H. & Burkart, V. Chaperones may cause the focus of diabetes autoimmunity on distinct (pro)insulin peptides. J. Autoimmun. 105, 102304. https://doi.org/10.1016/j.jaut.2019.102304 (2019).
Purcell, A. W., Sechi, S. & DiLorenzo, T. P. The evolving landscape of autoantigen discovery and characterization in type 1 diabetes. Diabetes 68(5), 879–886 (2019).
Delong, T. et al. Pathogenic CD4 T cells in type 1 diabetes recognize epitopes formed by peptide fusion. Science 351(6274), 711–714 (2016).
Pagetta, A. et al. Characterization of immune complexes of idiotypic catalytic and anti-idiotypic inhibitory antibodies in plasma of type 1 diabetic subjects. Mol. Immunol. 44(11), 2870–2883 (2007).
Schmidt, B. Z. & Perlmutter, D. H. Grp78, Grp94, and Grp170 interact with alpha1-antitrypsin mutants that are retained in the endoplasmic reticulum. Am. J. Physiol. Gastrointest. Liver Physiol. 289(3), G444-455 (2005).
Sandler, M., Gemperli, B. M., Hanekom, C. & Kühn, S. H. Serum alpha 1-protease inhibitor in diabetes mellitus: Reduced concentration and impaired activity. Diabetes Res. Clin. Pract. 5(4), 249–255 (1988).
Finotti, P., Piccoli, A. & Carraro, P. Alteration of plasma proteinase-antiproteinase system in type 1 diabetic patients. Influence of sex and relationship with metabolic control. Diabetes Res. Clin. Pract. 18(1), 35–42 (1992).
Odom, D. T. et al. Control of pancreas and liver gene expression by HNF transcription factors. Science 303(5662), 1378–1381 (2004).
Haliyur, R. et al. Human islets expressing HNF1A variant have defective β cell transcriptional regulatory networks. J. Clin. Invest. 129(1), 246–251 (2019).
Ghiasi, S. M. et al. endoplasmic reticulum chaperone glucose-regulated protein 94 is essential for proinsulin handling. Diabetes 68(4), 747–760 (2019).
Trajkovski, M. et al. Regulation of insulin granule turnover in pancreatic beta-cells by cleaved ICA512. J. Biol. Chem. 283(48), 33719–33729 (2008).
Cianciaruso, C. et al. Primary human and rat β-cells release the intracellular autoantigens GAD65, IA-2, and proinsulin in exosomes together with cytokine-induced enhancers of immunity. Diabetes 66(2), 460–473 (2017).
Blum, J. S., Wearsch, P. A. & Cresswell, P. Pathways of antigen processing. Annu. Rev. Immunol. 31, 443–473 (2013).
Acknowledgements
The authors would like to thank Prof. Stefano Toppo of the Department of Molecular Medicine (University of Padua) for critical reading of the manuscript and helpful suggestions.
Funding
This research received no specific grant from funding agencies in the public, commercial or not-for-profit sectors.
Author information
Authors and Affiliations
Contributions
P.F. conceived the work, designed research studies, searched and analysed the data and wrote the manuscript. A.P. analysed and elaborated the data, contributed to the discussion and reviewed the manuscript. Both authors gave their approval of the final version of the manuscript. P.F. is the guarantor of this work and, as such, has full access to all the data in the study and takes responsibility for the integrity of the data and accuracy of the data analysis.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Finotti, P., Pagetta, A. A mutant α1antitrypsin in complex with heat shock proteins as the primary antigen in type 1 diabetes in silico investigation. Sci Rep 11, 3002 (2021). https://doi.org/10.1038/s41598-021-82730-2
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-021-82730-2
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.