What has zinc transporter 8 autoimmunity taught us about type 1 diabetes?
Zinc transporter 8 (ZnT8), a protein highly specific to pancreatic insulin-producing beta cells, is vital for the biosynthesis and secretion of insulin. ZnT8 autoantibodies (ZnT8A) are among the most recently discovered and least-characterised islet autoantibodies. In combination with autoantibodies to several other islet antigens, including insulin, ZnT8A help predict risk of future type 1 diabetes. Often, ZnT8A appear later in the pathogenic process leading to type 1 diabetes, suggesting that the antigen is recognised as part of the spreading, rather than the initial, autoimmune response. The development of autoantibodies to different forms of ZnT8 depends on the genotype of an individual for a polymorphic ZnT8 residue. This genetic variant is associated with susceptibility to type 2 but not type 1 diabetes. Levels of ZnT8A often fall rapidly after diagnosis while other islet autoantibodies can persist for many years. In this review, we consider the contribution made by ZnT8 to our understanding of type 1 diabetes over the past decade and what remains to be investigated in future research.
KeywordsAutoimmunity B and T cells Biomarker Epitope specificity Method development Pathogenesis Prediction Review SLC30A8 Type 1 diabetes ZnT8 autoantibodies
Islet antigen-2 autoantibodies
Luciferase immunoprecipitation system
Zinc transporter 8
ZnT8 arginine-specific autoantibodies
ZnT8 tryptophan-specific autoantibodies
Islet autoantibodies develop before the clinical onset of type 1 diabetes and are currently the principal biomarkers for identifying individuals at risk of the disease. Detection of multiple autoantibody positivity in childhood implies an 80% risk of developing diabetes within 15 years, although the rate of progressions varies from a few months to decades . In 2007, zinc transporter 8 (ZnT8) autoantibodies (ZnT8A) were characterised  and have since been recognised [1, 3] as one of the four major islet autoantibodies, together with GAD65 autoantibodies (GADA) , islet antigen-2 autoantibodies (IA-2A)  and insulin autoantibodies (IAA) .
A summary of important unanswered questions about the immune response to ZnT8 in type 1 diabetes
Field of study
B cell responses
T cell responses
Initiation of autoimmunity
How do HLA genotypes alter central tolerance for ZnT8?
Is the C-terminal always the target of ZnT8A?
What causes ZnT8 to become a target of autoimmunity?
Do SNPs in SLC30A8 influence the rate of progression in people at risk of type 1 diabetes?
Does epitope spreading occur across ZnT8?
Are ZnT8A preferentially lost in rapid progressors or do they appear less often?
The broad T cell response to ZnT8 suggests previous epitope spreading; when does this occur?
Do common SNPs modulate age of onset and what is the mechanism of this?
What is the frequency of ZnT8A in adult-onset T1D?
How can it help us distinguish subtypes of diabetes?
How does HLA-A*24 present ZnT8 peptide compared with HLA-A*2?
Does the SLC30A8 genotype predict long-term clinical outcomes following diagnosis?
Why are ZnT8A lost more rapidly compared with GADA or IA-2A?
Does this correlate with beta cell function?
Why are ZnT8 T cell responses lost over the first few years after diagnosis?
Is this because of rapid loss of antigen?
The structure and function of ZnT8
The development of the ZnT8A immune response
The structure of ZnT8 suggests that less than half of the protein is accessible to immune surveillance and this may explain why ZnT8 is a target of autoimmune attack in type 1 diabetes. The whole protein or fragmented peptides are most likely to be accessible to the immune system upon beta cell death, although following granule exocytosis, during insulin secretion, the luminal transmembrane domains are exposed extracellularly .
Initiation of ZnT8 autoimmunity
At diagnosis, IAA, IA-2A and GADA are associated with specific HLA class II alleles contained within type 1 diabetes risk genotypes [19, 20]. In contrast, ZnT8A have not shown consistent HLA class II associations , except with DQ6.4 in individuals with type 1 diabetes . They are, however, associated with high diabetes risk (i.e. DR3/DR4) in first-degree relatives. Epitope scanning identified more putative ZnT8 epitopes for HLA-DQ2 than for -DQ8 or -DQ6.4 . An intermediate binding epitope for HLA-DQ2 was predicted to bind W325 but not R325, suggesting that this may contribute to differences in central tolerance for ZnT8. However, regardless of DQ type, people with diabetes had a higher frequency of proinflammatory ZnT8-specific CD4+ T cells than age- and HLA-matched non-diabetic individuals .
The predisposing HLA class I A*24 allele was negatively associated with presence of ZnT8A at and before diagnosis when age at onset was accounted for [25, 26]. Studies of ZnT8 epitopes for CD8+ T cells have focused on the better-characterised HLA-A*2. There is evidence that CD8+ T cells recognise a range of ZnT8 peptides across the transmembrane/loop and C-terminal regions in individuals with diabetes [27, 28, 29]. This could suggest previous epitope spreading but may also reflect the technical challenge of detecting and characterising antigen-specific T cell responses ex vivo. The number and phenotype of ZnT8-specific T cells identified in people with and without diabetes appear to be similar but functional differences have been identified, such as greater ZnT8-stimulated IFN-γ secretion by isolated CD8+ T cells from people with diabetes . In the pancreases of people who had type 1 diabetes, compared with type 2 diabetes or no diabetes, more ZnT8-specific CD8+ T cells were present, suggesting that ZnT8-specific lymphocytes home to the pancreas.
ZnT8 as an autoantigen
Humoral islet autoimmunity can develop in children as young as 6 months of age with a peak incidence of seroconversion at 2–3 years and a probable second peak in puberty . The first autoantibodies to develop in young children are typically IAA and/or GADA and, therefore, insulin and GAD65 are often regarded as primary targets of autoimmunity. In contrast, IA-2A and ZnT8A are rarely the only islet autoantibodies identified at primary seroconversion; data for ZnT8A are, however, more limited [2, 14]. Although ZnT8A can be detected close to initiation of autoimmunity, ZnT8A and IA-2A typically arise later in prediabetes and are more common in adolescents at diagnosis. Similarly, in mice, transfer of ZnT8 C-terminal-specific CD4+ T cells only caused diabetes if insulitis was present. Additionally, ZnT8-specific CD4+ T cells were only found in pancreas and lymph nodes of mice in advanced disease , supporting the observation that antigen spreading to ZnT8 is characteristic of the developing autoimmune response.
Most of the mature ZnT8A responses recognise the C-terminal of the protein and only 10% recognise the N-terminal . Within the C-terminal, ZnT8A can be specific to amino acid 325 of ZnT8 (Fig. 1c) and this specificity is defined by the SLC30A8 polymorphism rs13266634 . Hence, individuals with the CC genotype (R325) rarely develop ZnT8 tryptophan-specific autoantibodies (ZnT8WA) and individuals with the TT genotype (W325) rarely develop ZnT8 arginine-specific autoantibodies (ZnT8RA). Competitive displacement experiments show that ZnT8A are truly specific for R325 or W325 . Thus, individuals respond to endogenous ZnT8 protein determined by their own genome. This has not been as easy to demonstrate for the other islet antigens because they lack an amino acid polymorphism with such an obvious influence on the autoantibody response. In several populations, ZnT8A appear to cross-react with a viral protein from Mycobacterium avium subsp. paratuberculosis (e.g. ) and around 50% of ZnT8A-positive individuals have antibodies recognising epitopes independent of amino acid 325 . Therefore, molecular mimicry could be a contributing factor in the initial response to ZnT8 but more work is needed to evaluate this.
For other islet autoantibodies, epitope spreading has been demonstrated to occur during progression, and identification of epitopes indicative of later stages of disease has improved the specificity of these markers [35, 36]. Investigations into ZnT8 epitope autoantibody responses have focused on samples taken close to disease onset. In addition to the well-characterised amino acid 325R/W polymorphism, a second SNP, rs16889462 (G/A) which encodes Q325, has also been identified but is present in <1% of Europeans. A conformational ZnT8A epitope (dependent on R332, E333, K336 and K340 but independent of R325)  was characterised by comparing human and mouse chimeric ZnT8 proteins. The importance of conformational epitopes to ZnT8A responses is also supported by the observation that linear 15 amino acid ZnT8 peptides were insufficient to displace ZnT8A . If epitopes of ZnT8A associated with higher risk of diabetes could be identified, these would aid prediction but the pattern of epitope-specific responses before diagnosis has not so far proved useful for risk stratification.
Predicting diabetes using ZnT8A
Wenzlau et al first described ZnT8A in 2007; analysis of 43 individuals at high genetic risk followed prospectively showed that ZnT8A could stratify risk in those positive for IAA, GADA or IA-2A alone . Subsequently, the prospective birth-cohort study BABYDIAB also showed that ZnT8A aid prediction of disease in the presence or absence of other islet autoantibodies . The much larger TrialNet dataset confirmed these findings and concluded that screening for ZnT8A should be included in prediction and prevention studies . Several studies have shown that combined testing for IA-2A and ZnT8A identifies relatives who progress rapidly to disease in the most cost-effective way [40, 41, 42]. The contribution of ZnT8A in predicting risk is likely age-dependent; as the benefit of adding ZnT8A in single-islet-autoantibody-positive individuals in the Diabetes Autoimmunity Study in the Young (DAISY) was found when onset was after age 6 years . In preselected relatives who were positive for other islet autoantibodies, ZnT8A only added to risk prediction in relatives who were older or who were at genetically lower risk . Within those who progress slowly to disease, ZnT8A are frequently detected . Most studies of risk stratification for type 1 diabetes have been conducted in relatives or individuals with high HLA class II risk. The added predictive value of ZnT8A in the general population has not been fully assessed but is ongoing (e.g. in the German Fr1da study ). Despite contradictions, current evidence suggests that screening for ZnT8A will provide additional information about risk, especially in older individuals in whom early autoimmune responses to insulin are waning.
Lessons learnt from ZnT8A measured at diagnosis
When ZnT8A were originally characterised, these autoantibodies increased the number of people identified as single- or multiple-antibody positive at diabetes onset. Of 133 individuals negative for IAA, GADA and IA-2A at diagnosis, 26% were positive for ZnT8A and 47% of 51 people positive for a single antibody (IAA, GADA or IA-2A) became multiple-antibody positive when ZnT8A were considered . Overall, several international studies have shown that around two-thirds of children are ZnT8A positive at diagnosis [32, 46]; depending on the age group considered, the prevalence is similar to IA-2A.
The importance of ZnT8A as a biomarker in adult disease has not been fully assessed. One study from Belgium suggested that only half of those diagnosed over the age of 20 years are positive for ZnT8A . This frequency is less than for GADA (65%) but comparable with IA-2A (45%) [47, 48]. The contribution of ZnT8A in populations defined as LADA is difficult to determine because the characteristics of the participants vary between studies. There is, however, evidence that screening for ZnT8A reduces the cost of discriminating monogenic diabetes .
ZnT8A also contributed to our observation that the natural history of humoral autoimmunity may be changing. In our UK-based study, carried out between 1985 and 2002 in individuals at disease onset, the prevalence of IA-2A and ZnT8A increased while that of GADA and IAA remained stable over time . This could indicate a shift towards a more aggressive form of disease, as IA-2A and ZnT8A tend to develop later in pathogenesis. During the same time period, the incidence of diabetes in younger children has increased and the proportion of probands with high genetic risk has reduced, suggesting an increase in environmental risk [50, 51]. Lower genetic risk alleles, such as DQ6.4, could therefore have become more common in people with diabetes and contributed to increased ZnT8A. A Danish study, covering a shorter time span, found no difference in prevalence of IA-2A or GADA in individuals at the time of diagnosis (1997–2005) but ZnT8A were not measured .
ZnT8A as a biomarker of insulin secretory capacity post diagnosis
Another suggestion from the initial 2007 paper was that ZnT8A may be correlated with loss of insulin secretion because of ZnT8’s beta cell specificity and co-localisation with insulin . Following initially successful pancreas transplant, ZnT8A have been associated with rapid onset of hyperglycaemia and eventual loss of graft function, making them a potentially important biomarker for predicting beta cell loss . Ongoing beta cell death is likely to drive levels of ZnT8A [2, 54], although low-level expression in non-functional beta cells or neighbouring alpha cells may also contribute. After diagnosis ZnT8A are lost more rapidly than GADA or IA-2A , while T cell responses, particularly to the transmembrane domain, are also lost within a few years post diagnosis . This is important because if ZnT8A, which are easier to measure than T cells, reflect insulin secretory capacity and autoreactivity, they may be useful biomarkers for assessing/monitoring the efficacy of clinical trials. In 2013, Ingemansson et al showed that in a group of 15–34 year olds with diabetes (71% type 1), high C-peptide (indicative of endogenous insulin secretion) at diagnosis correlated with prospective preservation of ZnT8A 5 years after diagnosis . However, another study in individuals with younger onset showed that high positivity for ZnT8A at diagnosis was associated with lower C-peptide levels and increased insulin dose within the first 2 years after diagnosis, despite similar levels at disease onset . In a minority of individuals where ZnT8A persist for decades after diagnosis, cross-sectional studies have also failed to reach a consensus about the relationship between ZnT8A and C-peptide [57, 58, 59]. These contrasting findings could be explained if age at diagnosis influences this association. The hypothesis that ZnT8A could be used as a biomarker for therapeutic effect is attractive but is not strongly supported by current literature.
What benefits could new assays for ZnT8A offer for prediction?
Alternative assays, such as the luciferase immunoprecipitation system (LIPS) (Fig. 2c) and ELISAs (Fig. 2d), have begun matching or even exceeding the specificity and sensitivity of RIAs for GADA and IA-2A [62, 63]. While ELISAs eliminate the requirement for radioisotopes, sample volume requirements greatly exceed those of RIAs (increased from 4 μl per test to 50 μl); LIPS assays offer similar performance to RIAs with potentially even lower sample volume requirements . This would facilitate the use of small volume samples such as capillary bleeds for screening young infants and the general population.
There are assays in development that may maintain the conformational structure of ZnT8. In 2017 Wan et al developed a novel assay using full-length ZnT8 (R325) in combination with proteoliposomes (Fig. 2e) . This assay has the potential to detect ZnT8A that recognise epitopes outside the C-terminal but the need to use the pGOLD platform restricts routine application of this assay due to costs. Future assay adaptations to investigate other epitopes, including post-translational protein modification, and that are validated on ‘at-risk’ individuals before disease onset may improve disease prediction and should continue to be investigated.
Measurement of ZnT8A is a cost-effective method by which to identify multiple-islet-autoantibody-positive individuals at increased risk of diabetes. This marker may be particularly useful in assigning risk from adolescence onwards. The study of ZnT8A has also given us vital evidence of a true autoimmune response and strengthens observations that the pathogenesis of the disease may change over generations. In the future, the influence of variants in SLC30A8 on the rate of progression to clinical type 1 diabetes should be clarified, especially in those with age of onset above 15 years. Why autoimmunity spreads to target ZnT8 later in the pathogenesis of the disease, why this response is lost rapidly after diagnosis and whether these changes are related to beta cell function are just some of the questions that still require clarification. Analysis of ZnT8A and T cell epitopes before diagnosis will improve our understanding of disease pathogenesis and provide better biomarkers of disease.
The authors would like to thank K. M. Gillespie and A. J. K. Williams (University of Bristol) for their encouragement, guidance and advice about this review.
Both authors were responsible for drafting the article and revising it critically for important intellectual content. Both authors approved the version to be published.
Work in the authors’ laboratories is supported by Diabetes UK and JDRF. CLW was awarded a Diabetes UK studentship (16/0005556) and AEL was awarded a Diabetes UK and JDRF RD Lawrence fellowship (18/0005778 and 3-APF-2018-591-A-N).
Duality of interest
The authors declare that there is no duality of interest associated with this manuscript.
- 11.Xu K, Zha M, Wu X et al (2011) Association between rs13266634 C/T polymorphisms of solute carrier family 30 member 8 (SLC30A8) and type 2 diabetes, impaired glucose tolerance, type 1 diabetes: a meta-analysis. Diabetes Res Clin Pract 91(2):195–202. https://doi.org/10.1016/j.diabres.2010.11.012 CrossRefGoogle Scholar
- 23.Delli AJ, Vaziri-Sani F, Lindblad B et al (2012) Zinc transporter 8 autoantibodies and their association with SLC30A8 and HLA-DQ genes differ between immigrant and Swedish patients with newly diagnosed type 1 diabetes in the Better Diabetes Diagnosis study. Diabetes 61(10):2556–2564. https://doi.org/10.2337/db11-1659 CrossRefGoogle Scholar
- 25.Long AE, Gillespie KM, Aitken RJ, Goode JC, Bingley PJ, Williams AJ (2013) Humoral responses to islet antigen-2 and zinc transporter 8 are attenuated in patients carrying HLA-A*24 alleles at the onset of type 1 diabetes. Diabetes 62(6):2067–2071. https://doi.org/10.2337/db12-1468 CrossRefGoogle Scholar
- 33.Skarstrand H, Krupinska E, Haataja TJ, Vaziri-Sani F, Lagerstedt JO, Lernmark A (2015) Zinc transporter 8 (ZnT8) autoantibody epitope specificity and affinity examined with recombinant ZnT8 variant proteins in specific ZnT8R and ZnT8W autoantibody-positive type 1 diabetes patients. Clin Exp Immunol 179(2):220–229. https://doi.org/10.1111/cei.12448 CrossRefGoogle Scholar
- 35.Naserke HE, Ziegler AG, Lampasona V, Bonifacio E (1998) Early development and spreading of autoantibodies to epitopes of IA-2 and their association with progression to type 1 diabetes. J Immunol 161:6963–6969Google Scholar
- 36.Williams AJ, Lampasona V, Wyatt R et al (2015) Reactivity to N-terminally truncated GAD65(96-585) identifies GAD autoantibodies that are more closely associated with diabetes progression in relatives of patients with type 1 diabetes. Diabetes 64(9):3247–3252. https://doi.org/10.2337/db14-1694 CrossRefGoogle Scholar
- 41.Gorus FK, Balti EV, Vermeulen I et al (2013) Screening for insulinoma antigen 2 and zinc transporter 8 autoantibodies: a cost-effective and age-independent strategy to identify rapid progressors to clinical onset among relatives of type 1 diabetic patients. Clin Exp Immunol 171(1):82–90. https://doi.org/10.1111/j.1365-2249.2012.04675.x CrossRefGoogle Scholar
- 42.De Grijse J, Asanghanwa M, Nouthe B et al (2010) Predictive power of screening for antibodies against insulinoma-associated protein 2 beta (IA-2β) and zinc transporter-8 to select first-degree relatives of type 1 diabetic patients with risk of rapid progression to clinical onset of the disease: implications for prevention trials. Diabetologia 53:517–524CrossRefGoogle Scholar
- 43.Long AE, Gooneratne AT, Rokni S, Williams AJ, Bingley PJ (2012) The role of autoantibodies to zinc transporter 8 in prediction of type 1 diabetes in relatives: lessons from the European Nicotinamide Diabetes Intervention Trial (ENDIT) cohort. J Clin Endocrinol Metab 97(2):632–637. https://doi.org/10.1210/jc.2011-1952 CrossRefGoogle Scholar
- 48.Vandewalle CL, Falorni A, Svanholm S, Lernmark A, Pipeleers DG, Gorus FK (1995) High diagnostic sensitivity of glutamate decarboxylase autoantibodies in insulin-dependent diabetes mellitus with clinical onset between age 20 and 40 years. The Belgian Diabetes Registry. J Clin Endocrinol Metab 80:846–851Google Scholar
- 55.Ingemansson S, Vaziri-Sani F, Lindblad U, Gudbjornsdottir S, Torn C, Diss-Study G (2013) Long-term sustained autoimmune response to beta cell specific zinc transporter (ZnT8, W, R, Q) in young adult patients with preserved beta cell function at diagnosis of diabetes. Autoimmunity 46:50–61CrossRefGoogle Scholar
- 56.Juusola M, Parkkola A, Harkonen T et al (2016) Positivity for zinc transporter 8 autoantibodies at diagnosis is subsequently associated with reduced β-cell function and higher exogenous insulin requirement in children and adolescents with type 1 diabetes. Diabetes Care 39(1):118–121. https://doi.org/10.2337/dc15-1027 CrossRefGoogle Scholar
- 60.Bonifacio E, Yu L, Williams AK et al (2010) Harmonization of glutamic acid decarboxylase and islet antigen-2 autoantibody assays for national institute of diabetes and digestive and kidney diseases consortia. J Clin Endocrinol Metab 95(7):3360–3367. https://doi.org/10.1210/jc.2010-0293 CrossRefGoogle Scholar
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.