HLA-A gene variation modulates residual function of the pancreatic β-cells in children with type 1 diabetes

Introduction

Type 1 diabetes is an autoimmune T-cell-mediated disorder resulting from selective destruction of pancreatic -cells and progressive decline in endogenous insulin secretion capacity. Both genetic and environmental factors are involved in disease process. Despite recent progress in our knowledge about the disease, cellular, and molecular events associated with the initiation and progression of the auto-immune reaction against -cells remain poorly understood [1].
In humans, as well as in animal models, type 1 diabetes has a multigenic basis. The Major Histocompatibility Complex (MHC) region located on the short arm of chromosome 6 contains the major genes predisposing to type 1 diabetes [2–4]. The HLA region acco-unts for approximately 50% heritability of T1DM [2].
Association of type 1 diabetes with HLA class I genes was first observed. However, higher relative risk associated with HLA class II DR/DQ compared to HLA-A and -B genes has led to the conclusion that the primary locus of susceptibility to type 1 diabetes is located within the class II region. In Caucasians, predisposition to type 1 diabetes is mostly associated with the DRB1*03-DQB1*0201 and/or DRB1*04-DQB1*0302 haplotypes, while the DRB1*15-DQB1*0602 haplotype confers a strong protection from the disease [4].
In our previous study of a Polish population, the highest risk for disease development was conferred by the DRB1*04-DQB1*0302 haplotype [5]. These genes, however, cannot completely explain the association between HLA and T1D development. Significant type 1 diabetes associations were observed at all class I HLA loci indicating that HLA class I alleles, in addition to and inde-pendently from HLA class II alleles, are associated with type 1 diabetes [6].
Recently, the well-established dogma about the origins of insulitis has changed as the solid evidence from histopathological studies such as DiViD or nPOD (7) confirmed that islet cell upregulation of HLA-I expression is a genuine pathological feature in type 1 diabetes in humans [8]. This “hyperexpression” of HLA-I antigens is critical for early disease progression, promoting the effective engagement of influent CD8+ cytotoxic T cells specific to defined islet antigen [8–10]. It was also discovered that islet cell hypere-xpression of HLA-I can persist, at the protein level, beyond the initial phases of the disease [9]. This indicates the importance of some HLA class I alleles in T1D susceptibility and raises essential questions about the role of this phenomenon in disease progression and clinical presentation of the disease in humans.
A present study aimed to analyze the HLA-A gene association with a genetic predisposition to type 1 diabetes development and to evaluate the association of HLA class I and class II alleles with -cell destruction.

Material and methods

Patients

The study was conducted in a group of 108 unrelated type 1 diabetes children (51 girls and 57 boys) admitted to the Department of Pediatrics, Medical University of Lodz in Poland for initial treatment of type 1 diabetes. Their median age was 10.5 years (range: 1.0–17.0 years). All children were hyperglycaemic at the time of blood sampling. Blood glucose value ranged 160–998 mg/dl (mean 433 ±112 mg/dl).
The control group consisted of 92 unrelated, healthy blood donors from Central Poland.
All subjects gave their informed consent for inclusion before they participated in the study. The study was conducted in accor-dance with the Declaration of Helsinki, and the protocol was approved by the Bioethics Committee of Medical University of Lodz.

HLA typing

All subjects were typed for HLA-A, HLA-DRB1, and -DQB1 genes. Genomic DNA was extracted from peripheral blood cells using a conventional salting-out procedure. Genotyping was performed using hybridization with sequence-specific oligonucleotide probes following amplification of the corresponding gene by the PCR with 52 probes that were used for HLA-A typing, 25 for DRB1 and 24 for DQB1 as described previously [5].

Assessment of residual b-cell function in type 1 diabetes patients

The residual -cell function at the clinical onset of the disease (“day 0”) was assessed by analysis of plasma C-peptide concentration in all patients. Blood samples were drawn before the first dose of insulin and after ten days of insulin therapy (“day 10”). Plasma C-peptide level was estimated by the radioimmunoassay method (CIS-BioInternational, France).

Statistical analysis

Yates’ corrected chi2 or Fisher’s exact tests were used to compute-two-tailed p values, which were considered significant if less than 0.05. The power of association with the predisposition and clinical parameters of the disease with 95% confidence intervals (95% CI) was calculated using odds ratios (OR) according to Woolf’s formula.
An association between age, C-peptide concentration, and genetic factors was estimated by a t-test or Mann-Whitney U test ac-cording to the normality of data set. Comparison between “day 0” and “day 10” C-peptide levels was performed using the Wilcoxon signed-rank test. Correlation between C-peptide levels was assessed using the Spearman correlation test.
Pc indicates a p-value corrected by use of the Bonferonni inequality method, by multiplying p by the number of alleles compa-red.

Results

Association of HLA genes with type 1 diabetes

The distribution of HLA-A alleles among healthy controls corresponds to expected frequencies in the Caucasian population (Table I). 14 HLA-A alleles were found in our population. The most frequent allele in both control and type 1 diabetes groups was A*02 (33.2% and 33.7%, respectively). The second most frequent allele was A*03 in controls (17.4%) and A*25 in patients (12.0%). However, these differences were not statistically significant. The distribution of HLA-DRB1 and DQB1 alleles in our Polish population has previously been described [5].

Residual β-cell function in diabetic subjects

Evaluation of the residual -cell function was assessed by plasma C-peptide concentration at the clinical onset of the disease. A strong positive correlation between serum C-peptide levels at “day 0” and “day 10” (r = 0.71, p < 10-5) was found, with a median con-centration of 0.21 pmol/ml at day 0 and 0.19 at day 10 (p = 0.0004). This observation confirmed that this parameter is useful for evalua-tion of  cell function at disease-onset.
The C-peptide level was analyzed in patients by comparison with the lowest value of the normal range in healthy individuals (0.28 pmol/ml). Among diabetic children, 35% had C-peptide above this level at the time of diagnosis. Median C-peptide concentration in this “high” C-peptide group was significantly higher than in the “low” C-peptide group (0.47 pmol/ml, range: 0.36–0.79 and 0.16 pmol/ml, range: 0.11–0.21, respectively, p = 0.006). The potential relationship between age and C-peptide concentration was investiga-ted using the Spearman correlation test. Older children had a higher C-peptide level at the clinical onset (r = 0.4, p < 10–5).
To investigate the effect of HLA genes on beta-cell destruction, we analyzed the distribution of HLA-A, -DR, and -DQ alleles among “high” and “low” C-peptide diabetic children. No difference in DRB1 or DQB1 allele frequency was observed (data not shown). By contrast, analysis of HLA-A allele distribution revealed significant differences (Table II). Among “low C-peptide” individuals, HLA-A*02 frequency was 41.3 %, compared to 19.7 % among “high C-peptide” patients (Pc = 0.008, OR = 1.4, 95% CI: 1.2–1.7). Conver-sely, the HLA-A*26 allele was only detected once in the “high C peptide” group (0.7 %) compared to a frequency of 10.5 % in the other group (Pc < 0.007, OR = 0.15, 95% CI: 0.02–0.9). Moreover, HLA-A*02/*02 and A*02/X children were more likely to have “low” C-peptide value at disease onset compared to those with a non-A*02/non-A*02 genotype (p = 0.008, OR = 1.6, 95% CI: 1.3–2.0 and p = 0.015, OR = 1.4, 95% CI: 1.1–1.9, respectively). Analysis of C-peptide concentration according to HLA-A phenotype showed that median C-peptide levels at day 0 were significantly lower in A02 phenotype patients than in non-A02 (0.17 pmol/ml, range: 0.12–0.29, and 0.26 pmol/ml, range: 0.17–0.45, respectively; p = 0.008). Conversely, the C-peptide level was higher in A*26-positive compared to A*26–negative patients (median: 0.40 and 0.20, respectively, p = 0.04) (Table III). Similar results were obtained for genotype analysis (Table IV). Children with A*02/*02 genotype had significantly lower C-peptide median (0.20 pmol/ml, range: 0.14–0.24) compared to those with non-*02/non-*02 genotype (0.26 pmol/ml, range: 0.17–0.45, p = 0.008). In one patient A*26/*26 homozygote C-peptide level was 0.48 pmol/ml. In another patient,
A*02/A*26 heterozygote, C-peptide concentration was 0.78 pmol/ml – far above the minimal value in healthy individuals.

Discussion

Particular HLA class II haplotypes DRB1*04/DQA1*03/DQB1*03:02 have long been related to the risk of developing type 1 dia-betes and/or severity of insulin deficiency in type 1 diabetes [11]. We have now investigated whether an HLA class I gene can modulate the presentation of type 1 diabetes. Our results indicate that some HLA-A alleles can affect -cell function at the clinical onset of the disease. HLA-A*26 phenotype was associated with higher, and A*02 with lower C-peptide levels at disease onset. The strongest -cell destruction was observed in A*02/*02 patients, whereas non-A*02/non-A*02 genotype was associated with the highest insulin secretion at disease-onset.
This finding is especially intriguing in the light of recent data showing that HLA class I alleles have a crucial role in insulitis. In the setting of type 1 diabetes (T1D), insulitis lesions are enriched for CD8+ T cells, which are held as the final mediators of islet de-struction [8]. Interestingly, nearly all beta cell peptides identified as the antigenic targets of CD8+ T cells in type 1 diabetes patients (eg. native proteins, epitopes of proinsulin, glutamic acid decarboxylase (GAD), insulinoma-associated protein-2 (IA-2), and islet-specific glucose-6-phosphatase catalytic subunit-related protein (IGRP), a highly immunoprevalent zinc transporter 8), are recognized in the context of HLA-A*0201 [12]. Moreover, hyperexpression of HLA class I on beta cells in T1D is associated with interferon response seen in tissues infected by viruses [13].
According to the molecular mimicry hypothesis, viral proteins could share a particular sequence with beta-cell proteins (exam-ples: coxsackie and GAD [14], the rubella virus and GAD [15], rotavirus and IA-2 [16]. This model would at least partially explain the role of environmental factors in the pathogenesis of the disease.
Furthermore, CD4 helper T cells infiltrating pancreatic islets can enhance CTL-mediated destruction of the islets [17, 18]. The modulating effect of HLA class I alleles (HLA-A*02 and HLA-A*26) could modify antigen presentation to CTL and thus affect the rapidity of disease progression.
Another interesting hypothesis emerges from the studies of the thymus. Incomplete central tolerance mechanisms allow the survival of an islet reactive CD8+ T cell repertoire, which can be primed in the presence of defective peripheral immunoregulation and/or a proinflammatory islet microenvironment to progress toward T1D.
The thymic presentation of self-antigens to T cells also occurs via HLA class I molecules. Therefore it is reasonable to think that specific T1D-predisposing HLA class I alleles expressed in the thymus may contribute to insufficient thymic presentation of autoan-tigens to T cells [19, 20]. It has been shown by Bulek et al. that weak interactions between a preproinsulin peptide and HLA-A2 lead to suboptimal presentation to the TCR of responding CD8+ T-cells, which may more easily survive thymic selection [21].
The influence of HLA-A alleles on C-peptide concentration has also been observed in the Japanese population. Nakanishi re-ported an association between the presence of A*24 and complete -cell destruction [22].
Interestingly, in the Japanese as well as in our Polish population, the most frequent HLA-A allele is the one conferring accelera-ted -cell destruction. This finding may support the alternative hypothesis that autoimmune diseases are a side effect of a natural selec-tion process of HLA alleles that confer survival advantage by more efficient protection from infectious diseases [23].
In our study, HLA-A allele distribution did not show any difference between diabetic and control subjects, suggesting that this locus has no direct predisposing effect on type 1 diabetes development. This observation may result from the limited number of the pa-tients studied. Honeyman and colleagues, based on analysis of families at risk for type 1 diabetes, showed an increased frequency of HLA-A*24 in relatives who developed diabetes compared to those who did not [24]. Kobayashi et al. also found a significant associa-tion between the presence of HLA-A*24 and Bw*54 alleles and the childhood-onset of the disease in Japanese patients [25].
More recently, significant type 1 diabetes associations were observed at all class I HLA loci indicating that HLA class I alleles, in addition to and independently from HLA class II alleles, are associated with type 1 diabetes [6]. Among predisposing alleles are B*5701, B*3906, A*2402, A*0201, B*1801, C*0501, while A*1101, A*3201, A*6601, B*0702, B*4403, B*3502, C*1601, and C*0401 seem to be protective. Some alleles, notably B*3906, appear to modulate the risk of all DRB1-DQA1-DQB1 haplotypes on which they reside, suggesting a class I effect that is independent of class II. Other class I, type 1 diabetes associations appear to be spe-cific to individual class II haplotypes. Some apparent associations (e.g., C*1601) could be attributed to strong LD to another class I susceptibility locus (B*4403). A combination of HLA-A24, -DQA1*03, and -DR9 contributes to the acute-onset and early complete beta-cell destruction, whereas HLA-DR2 has a protective effect against complete beta-cell loss in type 1 diabetes [26]. These data indica-te that HLA class I alleles, in addition to and independently from HLA class II alleles, are associated with type 1 diabetes [6].
In type 1 diabetes (T1D), autoreactive cytotoxic CD8+ T cells are implicated in the destruction of insulin-producing -cells. The HLA-B*3906 and HLA-A*2402 class I genes confer increased risk and promote early disease onset, suggesting that CD8+ T cells that recognize peptides presented by class I molecules on pancreatic -cells play a pivotal role in the autoimmune response [27]. Mo-reover, -cell expression of class II molecules suggests that -cells may interact directly with islet-infiltrating CD4+ T cells and may play an immunopathogenic role [28]. In our study, no significant difference was observed according to the age of patients with different genotypes. Furthermore, higher C-peptide levels were observed in older healthy children suggesting that insulin secretion level is age-dependant.
The association of some HLA-A alleles with -cell damage raises a question about the possible participation of this gene in the development of diabetic complications. Indeed, in the Japanese population, A*24 is associated with early development of diabetic retino-pathy. As discussed above, A*24 predisposes to -cell destruction, hence to deprivation of endogenous C-peptide. On the other hand, C-peptide is thought to have a beneficial effect on some clinical parameters. There are pieces of evidence that diabetic patients with residual -cell function are less prone to the development of diabetic microangiopathy [29], nephropathy [29], and neuropathy [30]. These biolo-gical effects do not depend on insulin level and do not influence patients’ glucose blood levels [31]. In our study, 35% of children had normal C-peptide level at disease-onset. According to our results, this group would represent patients with low -cell destruction or high insulin secretion capacity. Therefore, one could expect some clinical consequences, such as lower progression of the disease, lower ini-tial insulin requirement, and fewer diabetic complications than patients with significantly lower plasma C-peptide at diagnosis. Confir-ming this hypothesis would require long follow-up and/or analysis of patients with the adult form of autoimmune diabetes (LADA).
In conclusion, our results suggest a different role of HLA class I and class II alleles in type 1 diabetes pathogenesis. DRB1 and DQB1 genes are involved in predisposition to the disease but play probably a minor role in -cell damage. By contrast, specific HLA-A alleles influence a mechanism of pancreatic -cell destruction. Such observation may be considered in trials evaluating treatments to maintain some -cell function and induce remission at (or after) disease onset.

Acknowledgments

This study was supported by grants from the Polish State Committee for Scientific Research No 4PO5E 05516 and grant from the Medical University of Lodz.

References

1. Pugliese A. Autoreactive T cells in type 1 diabetes. J Clin Invest 2017; 127: 2881–2891. doi: 10.1172/JCI94549
2. Todd JA, Farrall M. Panning for gold: genome-wide scanning for linkage in type 1 diabetes. Hum Mol Genet 1996; 5 Spec. No:1443–1448.
3. Nejentsev S, Howson JM, Walker NM, et al. Localization of type 1 diabetes susceptibility to the MHC class I genes HLA-B and HLA-A. Nature 2007; 450: 887–892. doi: 10.1038/nature06406
4. Noble JA, Valdes AM, Cook M, et al. The role of HLA class II genes in insulin-dependent diabetes mellitus: molecular analy-sis of 180 Caucasian, multiplex families. Am J Hum Genet 1996; 59: 1134–1148.
5. Krokowski M, Bodalski J, Bratek A, et al. HLA class II-associated predisposition to insulin-dependent diabetes mellitus in a Polish population. Hum Immunol 1998; 59: 451–455. doi: 10.1016/s0198-8859(98)00036-6
6. Noble JA, Valdes AM, Varney MD, et al. HLA class I and genetic susceptibility to type 1 diabetes: results from the Type 1 Diabetes Genetics Consortium. Diabetes 2010; 59: 2972–2979. doi: 10.2337/db10-0699
7. Krogvold L, Edwin B, Buanes T, et al. Detection of a Low-Grade Enteroviral Infection in the Islets of Langerhans of Living Patients Newly Diagnosed With Type 1 Diabetes. Diabetes 2015; 64: 1682. doi: 10.2337/db14-1370
8. Richardson SJ, Rodriguez-Calvo T, Gerling IC, et al. Islet cell hyper-expression of HLA class I antigens: a defining feature in type 1 diabetes. Diabetologia 2016; 59: 2448–2258. doi: 10.1007/s00125-016-4067-4
9. Wong FS, Karttunen J, Dumont C, et al. Identification of an MHC class I-restricted autoantigen in type 1 diabetes by screening an organ-specific cDNA library. Nat Med 1999; 5: 1026–1031. doi: 10.1038/12465
10. Mikk ML, Heikkinen T, El-Amir MI, et al. The association of the HLA-A*24:02, B*39:01 and B*39:06 alleles with type 1 diabetes is restricted to specific HLA-DR/DQ haplotypes in Finns. HLA 2017; 89: 215–224. doi: 10.1111/tan.12967
11. Cucca F, Lampis R, Congia M, Angius E, Nutland S, Bain SC, et al. A correlation between the relative predisposition of MHC class II alleles to type 1 diabetes and the structure of their proteins. Hum Mol Genet 2001; 10: 2025–2037. doi: 10.1093/hmg/10.19.2025
12. Di Lorenzo TP, Peakman M, Roep BO. Translational mini-review series on type 1 diabetes: Systematic analysis of T cell epi-topes in autoimmune diabetes. Clin Exp Immunol 2007; 148: 1–16. doi: 10.1111/j.1365-2249.2006.03244.x
13. Pavlovic D, van de Winkel M, van der Auwera B, et al. Effect of interferon-gamma and glucose on major histocompatibility complex class I and class II expression by pancreatic beta- and non-beta-cells. J Clin Endocrinol Metab 1997; 82: 2329–2336. doi: 10.1210/jcem.82.7.4055
14. Klemetti P, Hyoty H, Roivainen M, et al. Relation between T-cell responses to glutamate decarboxylase and coxsackievirus B4 in patients with insulin-dependent diabetes mellitus. J Clin Virol 1999; 14: 95–105. doi: 10.1016/s1386-6532(99)00050-5
15. Ou D, Jonsen LA, Metzger DL, Tingle AJ. CD4+ and CD8+ T-cell clones from congenital rubella syndrome patients with IDDM recognize overlapping GAD65 protein epitopes. Implications for HLA class I and II allelic linkage to disease suscepti-bility. Hum Immunol 1999; 60: 652–664. doi: 10.1016/s0198-8859(99)00037-3
16. Honeyman MC, Coulson BS, Stone NL, Gellert SA, Goldwater PN, Steele CE, et al. Association between rotavirus infection and pancreatic islet autoimmunity in children at risk of developing type 1 diabetes. Diabetes 2000; 49: 1319–1324. doi: 10.2337/diabetes.49.8.1319
17. Schoenberger SP, Toes RE, van der Voort EI, et al. T-cell help for cytotoxic T lymphocytes is mediated by CD40-CD40L inte-ractions. Nature 1998; 393: 480–483. doi: 10.1038/31002
18. Feau S, Garcia Z, Arens R, et al. The CD4(+) T-cell help signal is transmitted from APC to CD8(+) T-cells via CD27-CD70 in-teractions. Nat Commun 2012; 3: 948. doi: 10.1038/ncomms1948
19. Mohan JF, Petzold SJ, Unanue ER. Register shifting of an insulin peptide-MHC complex allows diabetogenic T cells to escape thymic deletion. J Exp Med 2011; 208: 2375–2383. doi: 10.1084/jem.20111502
20. Ferris ST, Carrero JA, Unanue ER. Antigen presentation events during the initiation of autoimmune diabetes in the NOD mou-se. J Autoimmun 2016; 71: 19–25. doi: 10.1016/j.jaut.2016.03.007
21. Bulek AM, Cole DK, Skowera A, et al. Structural basis for the killing of human beta cells by CD8(+) T cells in type 1 diabetes. Nat Immunol 2012; 13: 283–289. doi: 10.1038/ni.2206.
22. Nakanishi K, Kobayashi T, Murase T, et al. Human leukocyte antigen-A24 and -DQA1*0301 in Japanese insulin-dependent diabetes mellitus: independent contributions to susceptibility to the disease and additive contributions to acceleration of beta-cell destruction. J Clin Endocrinol Metab 1999; 84: 3721–3725. doi: 10.1210/jcem.84.10.6045
23. Price P, Witt C, Allcock R, et al. The genetic basis for the association of the 8.1 ancestral haplotype (A1, B8, DR3) with multi-ple immunopathological diseases. Immunol Rev 1999; 167: 257–274. doi: 10.1111/j.1600-065x.1999.tb01398.x
24. Honeyman MC, Harrison LC, Drummond B, et al. Analysis of families at risk for insulin-dependent diabetes mellitus reveals that HLA antigens influence progression to clinical disease. Mol Med 1995; 1: 576–582.
25. Kobayashi T, Tamemoto K, Nakanishi K, et al. Immunogenetic and clinical characterization of slowly progressive IDDM. Diabetes Care 1993; 16: 780–788. doi: 10.2337/diacare.16.5.780
26. Nakanishi K, Inoko H. Combination of HLA-A24, -DQA1*03, and -DR9 contributes to acute-onset and early complete beta-cell destruction in type 1 diabetes: longitudinal study of residual beta-cell function. Diabetes 2006; 55: 1862–1868. doi: 10.2337/db05-1049
27. Sidney J, Vela JL, Friedrich D, et al. Low HLA binding of diabetes-associated CD8+ T-cell epitopes is increased by post trans-lational modifications. BMC Immunol 2018; 19: 12. doi: 10.1186/s12865-018-0250-3.
28. Russell MA, Redick SD, Blodgett DM, et al. HLA Class II Antigen Processing and Presentation Pathway Components Demon-strated by Transcriptome and Protein Analyses of Islet beta-Cells From Donors With Type 1 Diabetes. Diabetes 2019; 68: 988–1001. doi: 10.2337/db18-0686
29. Johansson BL, Linde B, Wahren J. Effects of C-peptide on blood flow, capillary diffusion capacity and glucose utilization in the exercising forearm of type 1 (insulin-dependent) diabetic patients. Diabetologia 1992; 35: 1151–1158. doi: 10.2337/db18-0686
30. Johansson BL, Borg K, Fernqvist-Forbes E, et al. C-peptide improves autonomic nerve function in IDDM patients. Diabetolo-gia 1996; 39: 687–695. doi: 10.1007/BF00418540
31. Ido Y, Vindigni A, Chang K, et al. Prevention of vascular and neural dysfunction in diabetic rats by C-peptide. Science 1997; 277: 563–566. doi: 10.1126/science.277.5325.563