Genetic evidence for efficacy of targeting IL-2, IL-6 and TYK2 signalling in the prevention of type 1 diabetes: a Mendelian randomisation study Syventävien opintojen opinnäyte Lääketieteellinen tiedekunta Tekijä: Emilia Kaiser 11.11.2025 Turku Turun yliopiston laatujärjestelmän mukaisesti tämän julkaisun alkuperäisyys on tarkastettu Turnitin OriginalityCheck -järjestelmällä. Syventävien opintojen opinnäyte Tutkinto-ohjelma, oppiaine: Lääketieteen lisensiaatin tutkinto-ohjelma, Biolääketieteen laitos, Integratiivinen fysiologia ja farmakologia, DIPP-tutkimus Tekijä: Emilia Kaiser Otsikko: Genetic evidence for efficacy of targeting IL-2, IL-6 and TYK2 signalling in the prevention of type 1 diabetes: a Mendelian randomisation study Ohjaaja: LT Jaakko Koskenniemi Sivumäärä: 15 sivua Päivämäärä: 11.11.2025 Avainsanat: IL-2, IL-6, ko-lokalisaatio, mendelistinen satunnaistaminen, TYK2, tyypin 1 diabetes Tiivistelmä tutkimuksen sisällöstä: Tavoitteet Tutkimuksen tarkoituksena oli selvittää, tukeeko geneettinen näyttö IL-2-, IL-6- ja TYK2- signaalireittien kohdentamisen tehokkuutta tyypin 1 diabeteksen ehkäisyssä. Menetelmät Tutkimuksessa hyödynnettiin laajoja genominlaajuisia assosiaatiotutkimuksia (GWAS), ko- lokalisaatiota ja Mendelististä satunnaistamista arvioimaan geenien ilmentymisen vaikutusta tyypin 1 diabeteksen riskiin. Ko-lokalisaatiota käytettiin arvioimaan, vaikuttavatko samat geneettiset variantit diabeteksen riskiin ja kohdegeenien ilmentymisen. Mendelistisen satunnaistamisen avulla mitattiin kohdegeenin ilmentymisen vaikutusta tyypin 1 diabeteksen riskiin. Yksi ala-analyysi rajattiin myös geenivariantteihin, jotka muuttivat kohdegeenin koodaamaa proteiinia. Tulokset Ko-lokalisaatio osoitti, että IL2RA-, IL6R- ja IL6ST-geenien ilmentymiseen vaikuttavat geneettiset variantit lisäävät myös tyypin 1 diabeteksen riskiä. Mendelistisen satunnaistamisen perusteella IL2RA:n ilmentymisen lisääntyminen liittyi alhaisempaan diabeteksen riskiin, kun taas IL6R:n ja IL6ST:n ilmentymisen lisääntyminen lisäsi riskiä. Lisäksi TYK2-geenin funktionaaliset variantit osoittivat, että sen ilmentymätason väheneminen liittyy alhaisempaan riskiin sairastua tyypin 1 diabetekseen. Johtopäätökset ja tulkinta Tulokset tukevat näiden signaalireittien kohdentamisen potentiaalia tyypin 1 diabeteksen ehkäisyssä. Erityisesti IL-2:n modulointi ja IL-6:n sekä TYK2:n ilmentymistason säätely näyttäisivät olevan lupaavia strategioita, jotka voivat ohjata tulevia kliinisiä tutkimuksia ja ennaltaehkäisevien hoitomuotojen kehittämistä. IL-6-signalointi tyypin 1 diabeteksessa Emilia Kaiser Interleukiini 6 (IL-6) on keskeinen sytokiini, joka osallistuu laaja-alaisesti tulehduksen, immuunivasteen, aineenvaihdunnan ja hematopoieesin säätelyyn. Infektioiden ja kudosvaurion laukaisema makrofagien nopea IL-6:n eritys käynnistää ja tehostaa useita akuutin vaiheen vasteita [1]. IL-6:n on osoitettu estävän säätelijä-T-solujen (Treg) kehitystä ja toimintaa sekä lisäävän patogeenisten T-auttaja 17-solujen (Th17) erilaistumista [4]. Korostunut IL-6-signalointi voi siten horjuttaa Treg/Th17-suhdetta ja tämän häiriön on ehdotettu myötävaikuttavan autoimmuunisairauksien, kuten nivelreuman ja mahdollisesti myös tyypin 1 diabeteksen syntyyn [4,5]. IL-6 vaikuttaa sitoutumalla reseptorijärjestelmään ja aktivoi näin solunsisäisiä signalointireittejä. IL-6 sitoutuu reseptoriin, jota esiintyy kahdessa muodossa: solukalvoon kiinnittyneenä reseptorina (IL-6R) ja liukoisena muotona (sIL-6R). Näiden lisäksi IL-6-signalointi edellyttää gp130-proteiinia (IL6ST), joka toimii varsinaisena signalointialayksikkönä ja välittää solunsisäisen vasteen [2]. Neljän proteiinin muodostama IL-6-järjestelmä (IL-6, sen reseptori IL-6R, liukoinen reseptorimuoto sIL-6R ja signaalinvälittäjä gp130 eli IL6ST) mahdollistaa useita erilaisia signalointireittejä. IL-6- signalointi voi tapahtua klassisen reitin kautta, kun IL-6 sitoutuu solukalvon IL-6R:ään, tai trans- signalointina, jossa IL-6 sitoutuu liukoiseen IL-6R:ään (sIL-6R) ja muodostaa kompleksin solukalvoon sitoutuneen gp130:n kanssa. Lisäksi signalointi voi tapahtua trans-presentaation kautta, jossa dendriittisolujen pinnalla oleva sIL-6R sitoutuu T-solujen gp130-reseptoriin [2]. IL-6:n trans- signaloinnin ja trans-presentaation on ehdotettu olevan klassista signalointia voimakkaampia autoimmuunivasteen käynnistäjiä [3]. Lisäksi trans-signaloinnin estäminen (esim. sIL-6R:n salpaaminen) vaikuttaa lupaavalta, koska se voi vähentää tulehdusta ilman klassisen signaloinnin suojaroolin heikentämistä [2]. Kreikkalais-serbialaisessa havainnoivassa tutkimuksessa seurattiin IL-6-tasoja tyypin 1 diabetesta sairastavilla, ja ne olivat merkittävästi koholla henkilöillä, joiden diabetes oli huonosti hallinnassa iästä ja painoindeksistä riippumatta [6]. Tämä viittaa siihen, että heikko diabeteksen hoitotasapaino lisää systeemistä tulehdusvastetta autoimmuunidiabeteksessa, ja tukee IL-6:n käyttöä tulehduskuorman biomarkkerina [6]. Hundhausen et al. tutkivat IL-6-reitin merkitystä tyypin 1 diabeteksessa [7]. Tutkimuksessa havaittiin, että tyypin 1 diabetesta sairastavilla henkilöillä CD4- ja CD8-positiivisten T-solujen STAT1- ja STAT3-vasteet IL-6:lle olivat korostuneet. Lisäksi STAT3- positiivisten solujen määrä laski sairauden kestoon liittyen, mikä viittaa siihen, että IL-6-signalointi heikkenee pitkäaikaisessa tyypin 1 diabeteksessa. Tulokset osoittavat, että potilaiden T-solut ovat hyperreaktiivisia IL-6:lle [7]. Kliiniset tutkimukset IL-6 tai IL-6R signaloinnin kohdistuvista hoidoista tyypin 1 diabeteksen estämisessä ovat harvassa. Speake et al. tutkivat lyhytaikaisen IL-6- (siltuksimabi) tai IL-6R- (tosilitsumabi) estolääkityksen vaikutusta T-solujen aktivaatioon tyypin 1 diabetes -potilailla. Tosilitsumabi estää kaikkia kolmea IL-6-siganlointireittejä. Täydellinen IL-6R-estäminen ei kuitenkaan merkittävästi hidastanut beetasolukatoa tyypin 1 diabeteksen edetessä, mikä viittaa siihen, että pelkkä klassisen IL-6 signaloinnin estäminen ei vaikuta olevan tehokasta tyypin 1 diabeteksen etenemisen estossa. Siksi katse on kohdistunut tarkemmin kohdennettuihin lähestymistapoihin, kuten trans-signaloinnin estoon, voivat olla lupaavampia terapeuttisia hyöty-riskitasapainon vaihtoehtoja [8]. Greenbaum et al. tekemässä satunnaistetussa, lumekontrolloidussa ja kaksoissokkoutetussa kliinisessä tutkimuksessa havaittiin, että tosilitsumabi ei merkitsevästi hidastanut jäännösbeetasolutoiminnan heikkenemistä henkilöillä, joilla oli jo todettu tyypin 1 diabetes [9]. IL-6-signaloinnin monimuotoisuus ja sen vaikutus immuunivasteisiin sekä beetasolujen toimintaan tekevät siitä keskeisen tekijän tyypin 1 diabeteksen patogeneesissä. Klassisen, trans- ja trans- presentaatioreittien erilaisten vaikutusten ymmärtäminen on tärkeää, ja saattaa selittää, miksi yleinen IL-6-signaloinnin esto ei välttämättä ole tehokasta tai turvallista taudin etenemisen hidastamisessa. Kliinisissä tutkimuksissa hoidon kohdentaminen on ollut haastavaa, sillä vaikka IL-6:n eri signalointireittejä voidaan estää monin tavoin (vasta-aineet, JAK-inhibiittorit), patogeenisen tulehduksen vähentäminen ilman immuunipuolustuksen tai homeostaattisten toimintojen heikentymistä on vaikeaa [2]. Lisäksi optimaalinen ajankohta IL-6-kohdistuvalle hoidolle tyypin 1 diabeteksen ehkäisyssä on edelleen epäselvä. Tämä voi olla syynä Greenbaum et al. tutkimuksessa saatuihin tuloksiin [9], kun IL-6R-salpaus annettiin vasta taudin jo edettyä diagnoosiin. Meidän tutkimuksemme [10] osoitti, että IL6R- ja IL6ST-geenien ekspressiotason vaihtelu on kausaalisesti yhteydessä tyypin 1 diabeteksen riskiin. Tämä viittaa siihen, että IL-6-signalointi omalta osin edistää tyypin 1 diabeteksen etenemistä ja perustelee IL-6-signaloinnin tutkimista. 1. Tanaka T, Narazaki M, Kishimoto T. IL-6 in inflammation, immunity, and disease. Cold Spring Harb Perspect Biol. 2014;6(10):a016295. doi:10.1101/cshperspect.a016295 2. Garbers C, Heink S, Korn T, Rose-John S. Interleukin-6: designing specific therapeutics for a complex cytokine. Nat Rev Drug Discov. 2018;17(6):395–412. doi:10.1038/nrd.2018.45 3. Rose-John S, Jenkins BJ, Garbers C, Moll JM, Scheller J. Targeting IL-6 trans-signalling: past, present and future prospects. Nat Rev Immunol. 2023;23(10):666–681. doi:10.1038/s41577-023- 00856-y 4. Kimura A, Kishimoto T. IL-6: regulator of Treg/Th17 balance. Eur J Immunol. 2010;40(7):1830–1835. doi:10.1002/eji.201040391 5. Spence A, Tang Q. Restoring regulatory T cells in type 1 diabetes. Curr Diab Rep. 2016;16(11):110. doi:10.1007/s11892-016-0807-6 6. Koufakis T, Kouroupis D, Kourti A, et al. Interleukin-6-related inflammatory burden in type 1 diabetes: evidence for elevation with suboptimal glycemic control. J Clin Med. 2025;14(18):6511. doi:10.3390/jcm14186511 7. Hundhausen C, Roth A, Whalen E, et al. Enhanced T cell responses to IL-6 in type 1 diabetes are associated with early clinical disease and increased IL-6 receptor expression. Sci Transl Med. 2016;8(356):356ra119. doi:10.1126/scitranslmed.aad9943 8. Speake C, Habib T, Lambert K, et al. IL-6-targeted therapies to block the cytokine or its receptor drive distinct alterations in T cell function. JCI Insight. 2022;7(22):e159436. doi:10.1172/jci.insight.159436 9. Greenbaum CJ, Serti E, Lambert K, et al. IL-6 receptor blockade does not slow β-cell loss in new-onset type 1 diabetes. JCI Insight. 2021;6(21):e150074. doi:10.1172/jci.insight.150074 10. Heikkilä TE, Kaiser EK, Lin J, Gill D, Koskenniemi JJ, Karhunen V. Genetic evidence for efficacy of targeting IL-2, IL-6 and TYK2 signalling in the prevention of type 1 diabetes: a Mendelian randomisation study. Diabetologia. 2024;67(12):2667–2677. doi:10.1007/s00125- 024-06267-5 Vol.:(0123456789) Diabetologia (2024) 67:2667–2677 https://doi.org/10.1007/s00125-024-06267-5 ARTICLE Genetic evidence for efficacy of targeting IL‑2, IL‑6 and TYK2 signalling in the prevention of type 1 diabetes: a Mendelian randomisation study Tea E. Heikkilä1  · Emilia K. Kaiser1  · Jake Lin2,3  · Dipender Gill4  · Jaakko J. Koskenniemi1,5  · Ville Karhunen6,7,8 Received: 19 February 2024 / Accepted: 11 July 2024 / Published online: 13 September 2024 © The Author(s) 2024 Abstract Aims/hypothesis We aimed to investigate the genetic evidence that supports the repurposing of drugs already licensed or in clinical phases of development for prevention of type 1 diabetes. Methods We obtained genome-wide association study summary statistics for the risk of type 1 diabetes, whole-blood gene expression and serum protein levels and investigated genetic polymorphisms near seven potential drug target genes. We used co-localisation to examine whether the same genetic variants that are associated with type 1 diabetes risk were also associated with the relevant drug target genetic proxies and used Mendelian randomisation to evaluate the direction and magnitude of the associations. Furthermore, we performed Mendelian randomisation analysis restricted to functional variants within the drug target genes. Results Co-localisation revealed that the blood expression levels of IL2RA (encoding IL-2 receptor subunit α [IL2RA]), IL6R (encoding IL-6 receptor [IL6R]) and IL6ST (encoding IL-6 cytokine family signal transducer [IL6ST]) shared the same causal variant with type 1 diabetes liability near the corresponding genes (posterior probabilities 100%, 96.5% and 97.0%, respectively). The OR (95% CI) of type 1 diabetes per 1-SD increase in the genetically proxied gene expression of IL2RA, IL6R and IL6ST were 0.22 (0.17, 0.27), 1.98 (1.48, 2.65) and 1.90 (1.45, 2.48), respectively. Using missense variants, genetically proxied TYK2 (encoding tyrosine kinase 2) expression levels were associated with type 1 diabetes risk (OR 0.61 [95% CI 0.54, 0.69]). Conclusions/interpretation Our findings support the targeting of IL-2, IL-6 and TYK2 signalling in prevention of type 1 diabetes. Data availability The analysis code is available at https:// github. com/ jkosk ennie mi/ T1DSC REEN, which also includes instructions on how to download the original GWAS summary statistics. Keywords Co-localisation · IL-2 · IL-6 · Mendelian randomisation · TYK2 · Type 1 diabetes Abbreviations eQTL Expression quantitative trait locus / loci GWAS Genome-wide association study IL2RA IL-2 receptor subunit α IL2RB IL-2 receptor subunit β IL2RG IL-2 receptor subunit γ IL-6R IL-6 receptor IL6ST IL-6 cytokine family signal transducer JAK1 Janus kinase 1 JAK2 Janus kinase 2 JAK3 Janus kinase 3 LD Linkage disequilibrium MAF Minor allele frequency NK Natural killers pQTL Protein quantitative trait locus / loci Th17 T helper 17 Treg Regulatory T cell TYK2 Tyrosine kinase 2 Tea E. Heikkilä and Emilia K. Kaiser contributed equally to the manuscript and are joint first authors. Jaakko J. Koskenniemi and Ville Karhunen contributed equally to the manuscript and share senior authorship. Extended author information available on the last page of the article 2668 Diabetologia (2024) 67:2667–2677 Introduction Type 1 diabetes is an autoimmune disease characterised by the loss of beta cell function. Despite advances in continu- ous glucose monitoring and insulin administration, manag- ing type 1 diabetes remains a significant burden and few individuals reach glycaemic targets [1]. Several drugs have shown potential in delaying the loss in beta cell function in newly diagnosed diabetes [2] and teplizumab, a monoclonal anti-CD3 antibody, even delayed the onset of the clinical dis- ease by a median of 2 years [3]. However, no current therapy can completely halt the disease progression and therefore it is crucial to identify new efficacious drug targets. Drug targets backed by genetic evidence have higher success rates in clinical development [4]. Genome-wide association studies (GWAS) enable the discovery of genomic regions strongly associated with the disease of interest. These associations can be considered as evi- dence for the involvement of the corresponding proteins in the disease pathogenesis, implying that these proteins are potential drug targets for the disease. The evidence for a likely drug target can be further investigated via co- localisation and Mendelian randomisation. Co-localisation can be used to study whether the same causal variant is shared between the drug target and the disease liability or whether the loci of the risk allele and an allele influencing the drug target are distinct [5]. Furthermore, Mendelian randomisation can be used to assess how much the genetic variability in drug target levels affects the risk of a disease in the population [6]. To inform the prioritisation of targets for prevention of type 1 diabetes, we aimed to investigate the genetic evidence for the efficacy of 12 drug targets in prevention of type 1 diabetes. These targets were selected because a prior GWAS reported that they are associated with the risk of type 1 diabetes and drugs that target them are already licensed for indications other than type 1 diabetes or have progressed to clinical development (see electronic sup- plementary material [ESM] Table 1). Methods Study design We selected the drug targets based on a previous GWAS of type 1 diabetes by Robertson et al [7] (Table 1 and ESM Fig. 1). Using a priority index, we ranked the drugs based on the following four factors: (1) existence of genetic variant(s) close to the potential target; (2) chromatin accessibility; (3) gene expression data in relevant cell types; and (4) protein–protein interactions [8]. We focused on 12 proteins (IL-2 receptor subunit α [IL2RA], IL-2 receptor subunit β [IL2RB], IL-2 receptor subunit γ [IL2RG], IL-6 receptor [IL6R], IL-6 cytokine family signal transducer [IL6ST], IL-12 subunit β [IL12B], 2669Diabetologia (2024) 67:2667–2677 IL-23 subunit α [IL23A], IFN-α and -β receptor subunit 2 [IFNAR2], Janus kinase 1 [JAK1], Janus kinase 2 [JAK2], Janus kinase 3 [JAK3] and tyrosine kinase 2 [TYK2]) that have already been targeted in clinical trials for autoim- mune diseases and some of which have been licensed for indications other than type 1 diabetes (ESM Table 1) [7]. Among these targets, eight target genes (IL2RA, IL2RB, IL6R, IL6ST, IL23A, JAK2, JAK3, TYK2) had a locus asso- ciated with the risk of type 1 diabetes (p<1 ×  10−5) within a distance of 1 million base pairs from the target gene (see ESM Fig. 2). IL23A was excluded from the analy- ses because data on its circulating levels were unavail- able, leaving seven targets for the subsequent analyses. Genomic regions under investigation are listed in Table 1. All primary studies that generated the GWAS summary statistics used in our analysis have undergone institutional board review, have received ethical approval and were con- ducted according to declaration of Helsinki [9–11]. Participants We obtained the GWAS summary statistics for type 1 diabetes, whole-blood gene expression (expres- sion quantitative trait loci [eQTL]) and serum protein lev- els (protein quantitative trait loci [pQTL]) (ESM Table 2). eQTL data were used for loci near the drug target genes since they encode intracellular or membrane-bound proteins. However, pQTL data were also analysed for loci in the vicin- ity of IL6ST and IL6R genes, since soluble forms of their proteins (gp130 and IL-6 receptor [IL-6R], respectively) may also modulate IL-6 signalling (see ESM Fig. 3) [12]. We included loci within 1 million base pairs of the seven potential drug target genes (Table 1). We obtained the data on type 1 diabetes risk variants from a subsequent GWAS of 18,942 cases and 501,638 controls of European ancestry from nine cohorts [9]. We obtained pQTL data for IL6ST from a GWAS of 35,559 Icelanders and eQTL data from GWAS of 31,684 individuals from 37 eQTLGen Consortium cohorts, most individuals being European [10, 11]. We sum- marise the study population details of the utilised GWAS, as well as methods of ascertainment of cases of type 1 diabetes, measurement of whole-blood gene expression and serum protein levels, in ESM Methods. Co‑localisation We conducted co-localisation analysis to assess whether the genetic associations for type 1 diabetes risk near the seven drug target genes align with those for the whole-blood gene expression or serum protein levels of these targets. We performed a co-localisation analysis using a ‘coloc’ package in R [5]. This method uses Bayes- ian principles to assess the relationship between two traits. It considers all variants within a specific genetic locus and evaluates the following hypotheses (H0–H4), assuming a maximum of one causal variant per trait: H0: there is no association with either trait, implying no specific causal variants H1: there is an association with the exposure trait only H2: there is an association with the outcome trait only H3: there are associations with both exposure and out- come traits, two independent SNPs (i.e. distinct casual variants) H4: there are associations with both exposure and out- come traits, one shared causal variant A high posterior probability for H4 implies a shared causal variant for the two traits. A substantial posterior prob- ability for H3 suggests the two traits are influenced by dis- tinct causal variants linked to each trait. We used the default prior probabilities of 1 ×  10−4, 1 ×  10−4 and 1 ×  10−5 for a variant being associated with the exposure trait, the outcome trait and both traits, respectively. Table 1 Genomic regions under investigation a Using genome build hg38 b Within ±1 million base pairs from the gene, referring to eQTL data if not stated otherwise c Both pQTL and eQTL data were analysed because soluble and membrane-bound IL-6R and gp130 (encoded by IL6ST) have different biological actions (see ESM Fig. 3) Gene Chromosome Start positiona End positiona Variants availableb IL2RA 10 6,010,689 6,062,367 9683 IL2RB 22 37,118,666 37,175,118 5489 IL23A 12 56,338,884 56,340,410 No pQTL data available IL6Rc 1 154,405,193 154,469,450 pQTL:8116 eQTL: 4274 IL6STc 5 55,935,095 55,995,022 pQTL:12,024 eQTL:4253 JAK2 9 4,984,390 5,129,948 7939 JAK3 19 17,824,780 17,848,071 7142 TYK2 19 10,350,533 10,380,608 6403 2670 Diabetologia (2024) 67:2667–2677 Mendelian randomisation To evaluate the direction and magnitude of the causal effects, we performed Mendelian randomisation for those variants that co-localised between drug target levels and the risk of type 1 diabetes (posterior probability for H4 >0.8). Mendelian randomisation uses genetic variants to investigate the relationship between an exposure (e.g. drug target levels) and an outcome (e.g. type 1 diabetes risk) for causality. Under Mendel’s law of assort- ment, genetic variants are accepted to be independent of other genetic alleles and can be used as valid instrumental variables to estimate the causal effect of the exposure on the outcome if the following three assumptions are fulfilled [6]. 1: the genetic variant is associated with the exposure 2: the genetic variant is associated with the outcome only through the exposure 3: the genetic variant is not associated with any confound- ers Advantages of Mendelian randomisation include limited susceptibility to reverse causation and confounding by exter- nal factors that influence both exposure and the outcome. Linkage disequilibrium (LD) may confound Mendelian ran- domisation if genetic variants are not shared between the exposure and outcome but they reside in the same genomic area. However, we investigated this possibility in the prior co-localisation step. Examination of functional variants near drug target genes To further examine the potential causality of the putative drug targets on type 1 diabetes risk, we searched for functional missense variants in the coding area of these seven drug target genes from PhenoScanner [13] that were associated with the protein/expression levels of the target at p<1 ×  10−5. We sought for missense variants only within the gene region for each gene. We used a more lenient thresh- old than the genome-wide significance of p<5 ×  10−8 since we only focused on cis-variants, and the threshold used can be interpreted as a Bonferroni-corrected threshold for 5000 independent variants. These variants were then individually used as instruments in Mendelian randomisation to test for causality of the targets on the risk of type 1 diabetes. Tissue‑specific gene expression analyses using bulk tissue and single‑cell eQTL data To further assess the tissue speci- ficity of the regions that showed evidence for association in Mendelian randomisation, we conducted co-localisation with tissue-specific gene expression and type 1 diabetes lia- bility. We obtained data for IL2RA, IL6R, IL6ST and TYK2 single-cell eQTL of data for immune cell subsets from a GWAS of 982 individuals from the OneK1K cohort, and spleen and pancreas eQTL data from the Genotype Tissue Expression (GTEx) project (v8) [14, 15]. Statistical analysis All analyses were done with R version 4.2 (R Foundation for Statistical Computing, Vienna, Austria) using the ‘coloc’ and ‘TwoSampleMR’ packages [5, 16]. Effect size for change in the risk of type 1 diabetes is reported per change in SD of serum protein or mRNA levels, and the genetic variants were not weighted in any of the analyses. β and SE values for whole-blood RNA levels were calculated using for- mulae 훽 = z (√ 2p(1 − p) ( n + z2 ))−1 and SE = 2p(1 − p)(n + z2)−1 , where p = minor allele frequency, n = sample size, and z = z score. Mendelian randomisation estimates are reported as Wald estimates. To assess the instrument strength and potential weak instrument bias, we calculated the F statistics for the instru- ments using formula F = (β/SE)2. We only used variants that were available in the GWAS summary statistics for both exposure and outcome traits within each genomic locus. All the LD r2 values reported in this study were obtained from the European population of the 1000 Genomes project using the package ‘ieugwasr’. The study protocol was not preregistered for this study. The code used to generate our results is available at https:// github. com/ jkosk ennie mi/ T1DSC REEN. Results The same genetic variants affect type 1 diabetes risk and whole‑blood IL2RA, IL6R and IL6ST gene expression We found evidence for a shared causal variant between the risk of type 1 diabetes and whole-blood IL2RA (rs61839660, pos- terior probability 100%, Fig. 1), IL6R (rs10908839, poste- rior probability 96.5%, Fig. 2) and IL6ST gene expression (rs7731626, posterior probability 97.0%, Table 2 and ESM Fig. 4). The variant rs10908839 is in strong LD (r2=0.76) with the previously identified lead SNP rs2229238 associ- ated with the risk of type 1 diabetes in the study by Robert- son et al [7]. No evidence for co-localisation was observed between drug target levels and the risk of type 1 diabetes near the other target genes (Table 2 and ESM Figs 5–10). Genetically proxied variability in IL2RA, IL6R and IL6ST expression is associated with the risk of type 1 diabetes We investigated, using Mendelian randomisation, the direction and magnitude of the expected change in the risk of type 1 diabetes if IL2RA, IL6R and IL6ST were targeted, using vari- ants identified as the most likely shared causal variant in co- localisation (rs61839660 for IL2RA gene expression, F=248; variant rs10908839 for IL6R gene expression, F=148; and variant rs7731626 for IL6ST expression, F=159). The results revealed an OR of 0.22 (95% CI 0.17, 0.27) for type 1 diabetes risk (p=5.3 ×  10−43) per SD increase in genetically proxied IL2RA expression, an OR of 1.98 (95% CI 1.48, 2671Diabetologia (2024) 67:2667–2677 2.65) per SD increase in genetically proxied IL6R expres- sion (p=5.2 ×  10−6) and an OR of 1.90 (95% CI 1.45, 2.48) per SD increase in genetically proxied IL6ST expression (p=2.6 ×  10−6; Table 3). Functional variant in coding area of TYK2 is associated with the risk of type 1 diabetes We found 22 missense mutations in the examined drug target genes with GWAS summary data on type 1 diabetes risk (ESM Table 3). Most of the mutations had an allele frequency of <0.01 with the nota- ble exceptions of rs2228145 (minor allele frequency [MAF] =0.39) in IL6R as well as rs12720356 (MAF=0.09) and rs2304256 (MAF=0.28) in TYK2. Summary statistics for both drug target level and type 1 diabetes risk were avail- able for nine variants. One independent variant at LD r2<0.2 within each locus (rs41316003 in JAK2, rs2304256 rs61839660 0 10 20 30 40 50 60 5800 5900 6000 6100 6200 − lo g 1 0 (p v a lu e ) a 0 10 20 30 40 50 5800 5900 6000 6100 6200 − lo g 1 0 (p v a lu e ) b ANKRD16 FBH1 GDI2 IL15RA IL2RA PFKFB3 RBM17 5800 5900 6000 6100 6200 Chromosome 10 position (kB) c r 2 with rs61839660 0.8−1.0 0.6−0.8 0.4−0.6 0.2−0.4 0.0−0.2 rs61839660 Fig. 1 Regional Manhattan plot of whole-blood IL2RA expression (a) and risk of type 1 diabetes (b) near IL2RA (c). The colours indicate the LD r2 value with rs61839660 (the most likely shared causal variant identified in co-localisation), based on 1000Genomes European reference   2672 Diabetologia (2024) 67:2667–2677 in TYK2 and rs141500365 IL6ST) was associated with the relevant drug target (protein or gene expression levels at p<1 ×  10−5). When using these variants as instruments in Mendelian randomisation, the genetically proxied TYK2 expression was associated with the risk of type 1 diabetes (rs2304256: OR 0.61 [95% CI 0.54, 0.69]) whereas we found no clear evidence for association when using missense vari- ants in IL6ST (OR 0.98 [95% CI 0.78, 1.20]) or JAK2 (OR 0.74 [95% CI 0.46, 1.20]; Table 3). ESM Table 4 shows the LD matrix between the functional variants in the coding area of TYK2 as well as the lead SNPs associated with TYK2 expression and the risk of type 1 diabetes in the vicinity of the TYK2 gene. 0 5 10 15 20 25 30 35 154,300 154,400 154,500 154,600 154,700 − lo g 1 0 (p v a lu e ) a 0 1 2 3 4 5 6 154,300 154,400 154,500 154,600 154,700 − lo g 1 0 (p v a lu e ) b ADAR AQP10 ATP8B2 CHRNB2 IL6R SHE TDRD10 UBE2Q1 154,300 154,400 154,500 154,600 154,700 Chromosome 1 position (kB) c r 2 with rs10908839 0.8−1.0 0.6−0.8 0.4−0.6 0.2−0.4 0.0−0.2 rs10908839 rs10908839 Fig. 2 Regional Manhattan plot of whole-blood IL6R expression (a) and risk of type 1 diabetes (b) near IL6R (c). The colours indicate the LD r2 value with rs10908839 (the most likely shared causal variant identified in co-localisation), based on 1000Genomes European reference 2673Diabetologia (2024) 67:2667–2677 Type 1 diabetes and IL2RA co‑localise in CD8+ effector memory T cells and type 1 diabetes and IL6ST in CD4+ and CD8+ naive and central memory T cells To study which specific blood immune cells or tissues mediate the asso- ciation between IL2RA, IL6R, IL6ST and TYK2 expres- sion and type 1 diabetes risk, we analysed their pairwise co-localisation with type 1 diabetes risk in spleen, pancreas and subsets of blood immune cells. IL2RA expression in CD8+ effector memory T cells co-localised (posterior prob- ability 99.4%, ESM Table 5) with type 1 diabetes risk, with the lead causal variant rs61839660 being the same as for whole-blood IL2RA mRNA expression. Neither IL6R nor TYK2 expression co-localised in any of available immune cells. However, IL6ST expression co-localised in CD4+ and CD8+ naive and central memory T cells (lead causal variant rs7731626; posterior probabilities 97.0% and 92.4%, respectively, ESM Table 5). We did not observe such robust evidence for co-localisation between IL2RA, IL6R, IL6ST or TYK2 eQTL and the risk of type 1 diabetes in other cell types (ESM Table 5) or in spleen or pancreas (ESM Table 6). Discussion Leveraging data from large-scale GWAS and multiple quan- titative trait locus datasets, we investigated the genetic evi- dence for the efficacy of seven candidate drugs in preven- tion of type 1 diabetes. Using co-localisation and Mendelian randomisation, we found genetic evidence to support the role of IL-2 and IL-6 signalling in the pathogenesis of type 1 diabetes. In addition, the investigation of functional mis- sense variants suggested that TYK2 signalling is involved in the aetiology of type 1 diabetes. While the original GWAS of immune cell subsets did not report the eQTL of IL2RA in regulatory T cells (Tregs), our evidence for the protective effect of blood IL2RA expression on the risk of type 1 diabetes could be interpreted as sup- porting the role of Tregs as a natural protection against type 1 diabetes. A low but sufficient level of IL-2 is crucial for the survival and function of Tregs, which constitutively express IL2RA, IL2RB and IL2RG to produce α-, β- and γ-chains, respectively, required for trimeric high-affinity IL-2 recep- tors [17]. Naive T cells express IL2RB and IL2RG, required Table 2 Results of co-localisation between drug target (eQTL/pQTL) and risk of type 1 diabetes Data are presented as posterior probabilities (%) H0, no association with either trait; H1, an association with the exposure trait only; H2, an association with the outcome trait only; H3, associations with both exposure and outcome traits, two independent SNPs (i.e. distinct causal variants); H4, associations with both exposure and outcome traits, one shared causal variant Gene Trait H0 H1 H2 H3 H4 IL2RA eQTL 0.0 0.0 0.0 0.0 100.0 IL2RB eQTL 0.0 0.0 0.0 99.7 0.3 IL6R eQTL 0.0 0.5 0.0 2.9 96.5 IL6R pQTL 0.0 46.1 0.0 51.7 2.2 IL6ST eQTL 0.0 0.3 0.0 2.8 97.0 IL6ST pQTL 0.0 9.3 0.0 90.5 0.2 JAK2 eQTL 0.0 0.0 0.0 100.0 0.0 JAK3 eQTL 0.0 1.2 0.3 98.5 0.0 TYK2 eQTL 0.0 0.0 0.0 86.6 13.4 Table 3 Mendelian randomisation results for IL2RA, IL6R, IL6ST, JAK2 and TYK2 Target Lead variant identified in co-localisation Mendelian randomisation OR (95% CI) p value Lead functional variant Mendelian randomisation OR (95% CI) p value IL2RA rs61839660 0.22 (0.17, 0.27) 5.3 ×  10−43 – – – IL6R rs10908839 1.98 (1.48, 2.65) 5.2 ×  10−6 – – – IL6ST rs7731626 1.90 (1.45, 2.48) 2.6 ×  10−6 rs141500365 0.98 (0.78, 1.20) 0.84 JAK2 – – – rs41316003 0.74 (0.46, 1.20) 0.22 TYK2 – – – rs2304256 0.61 (0.54, 0.69) 1.4 ×  10−14 2674 Diabetologia (2024) 67:2667–2677 for the intermediate-affinity IL-2-receptors, but they express IL2RA only transiently when stimulated by antigen-present- ing cells and are thus less stimulated by IL-2 when not acti- vated. Thus, higher blood IL2RA expression could be a sign of increased quantity and function of Tregs, which maintain a level of tolerance towards self-peptides and decrease the risk of type 1 diabetes [18]. A previous small study showed that rs12722495, which is in strong LD (r2=0.89) with our lead IL2RA eQTL and type 1 diabetes risk locus rs61839660 near IL2RA, decreased IL2RA expression in Tregs as well as their sensitivity to IL-2 [19]. Alternatively, as IL-2 signalling increases the prolifera- tion of the conventional T cells and Tregs alike, IL-2 sig- nalling might not decrease the risk of type 1 diabetes only by increasing tolerance but also by promoting appropriate responses to pathogens. This is supported by our single- cell-level results, in which we observed evidence for co- localisation between IL2RA expression and the risk of type 1 diabetes only in CD8+ central memory T cells (eQTL of Tregs were not available as they were not distinguished from other T cells). In a birth cohort study of children at high genetic risk of type 1 diabetes (TEDDY study), presence of enteroviral DNA in stool was associated with the risk of islet autoimmunity and the association was stronger in persistent infections indicated by prolonged shedding of enteroviral DNA [20]. Likewise, children who developed islet autoimmunity presented longitudinal transcriptional signatures consistent with a less-robust immune response against enteroviral infections compared with matched control children [21]. Since the increased IL-2 signalling in CD8+ T cells during viral infections prioritises robust immune response against production of long-lived memory cells, this explanation might also explain why rs61839660 near IL2RA strongly co-localised between the risk of type 1 diabetes and the eQTL of IL2RA in effector memory T cells and CD8+ naive/central memory T cells. The co-localisation of IL2RA expression in CD8+ effec- tor memory T cells with type 1 diabetes risk is further sup- ported by the previous reports of IL-2 impairment leading to CD8+ T cell exhaustion, potentially driven by both acute and chronic viral infections [22, 23]. Interestingly, our lead rs61839660 near IL2RA, which was associated with increased IL2RA expression and decreased risk of type 1 diabetes, was previously shown to be associated with higher risk of Crohn’s disease [24] and lower risk of type 1 diabetes [25]. This suggests that the optimal balance between effector and regulatory T cell function may vary between autoim- mune diseases [17]. Regardless of the possible mechanism, our findings support the rationale of conducting type 1 dia- betes prevention trials with low-dose IL-2. Consistent with our finding of the protective effect of IL-2 signalling, we found that IL-6 signalling increased the risk of type 1 diabetes. The secretion of IL-6 from macrophages rapidly in response to infections and tissue damage pro- motes various acute phase responses [26]. IL-6 signalling occurs as classic signalling through a cell-membrane- bound IL-6R, trans-signalling through soluble circulating IL6R and membrane-bound gp130 (encoded by IL6ST) and trans-presentation by dendritic cells, in which IL-6 is pre- sented to membrane-bound gp130 in T cells via dendritic cell-membrane-bound IL-6R (ESM Fig. 3) [27]. IL-6 inhib- its the development and function Tregs and promotes the development of pathogenic T helper 17 (Th17) cells [28], which is inhibited by IL-2. Pronounced IL-6 signalling may alter the balance of Treg/Th17, a proposed causative factor in autoimmune diseases such as rheumatoid arthritis [28] and possibly also type 1 diabetes [29]. While, IL-6 trans-signalling and trans-presentation may be a more potent inducer of autoimmunity than classic sig- nalling [12], our finding that blood IL6R and IL6ST expres- sion are associated with increased risk of type 1 diabetes may be explained by any of the three signalling modalities. However, since IL6ST expression in CD4+ and CD8+ naive or central memory T cells co-localised with type 1 diabetes and IL6R expression did not, it is tempting to speculate that trans-signalling might be more important than classic IL-6 signalling in the pathogenesis of type 1 diabetes. In contrast to our findings, tocilizumab (a monoclonal anti- body against IL6R), which blocks all three IL-6 signalling modalities, did not significantly affect the decline in residual beta cell function in individuals with newly diagnosed type 1 diabetes in a randomised, placebo-controlled, double-blind clinical trial [30]. However, this discrepancy may be partially explained by the timing of the intervention. Genetic polymor- phisms typically exert life-long influence on risk of diseases, including every stage of type 1 diabetes, whereas the interven- tion in the study by Greenbaum et al [30] took place after the diagnosis of type 1 diabetes, at which stage a sharp fall in beta cell function has already taken place [31]. Whole-blood TYK2 expression and the risk of type 1 dia- betes did not co-localise, whereas when using the missense mutation rs2304256 in TYK2 as an instrument in Mendelian randomisation, TYK2 expression was associated with type 1 diabetes risk. The absence of evidence for co-localisation may reflect violations of the one-causal-variant assumption. Indeed, the missense mutation rs2304256 is in very high LD (r2=1.00) with the lead TYK2 eQTL rs34725611. Moreover, rs2304256 is only in moderate LD (r2=0.10) with rs144309607, the lead variant on type 1 diabetes liability in co-localisation, suggest- ing two independent signals. Of note, a known missense vari- ant rs34536443 is not available in the eQTL data and therefore could not be used in co-localisation. Despite the one-causal- variant assumption, ‘coloc’ is relatively robust to multiple causal variants, and co-localisation methods allowing for mul- tiple causal variants are highly sensitive to LD misspecifica- tions in the reference panel [6]. Therefore, in the absence of an 2675Diabetologia (2024) 67:2667–2677 accurate LD reference, we proceeded with the ‘coloc’ method for our co-localisation while acknowledging its limitations. Overall, our TYK2 findings are consistent with previous studies suggesting that TYK2 signalling is associated with the risk of type 1 diabetes [32]. Promising results have been found in clinical trials targeting TYK2 in autoimmune dis- eases, supporting the potential of drug repurposing in type 1 diabetes [33]. Other reasons for the co-localisation discord- ance could be that TYK2 is not activated until IFN-α binding to IFNAR1 and TYK2 RNA expressions are known to have low tissue and cell type specificity [34, 35]. It is important to note that Mendelian randomisation estimates are only valid if the instrumental variable asso- ciations (relevance, independence, exclusion restriction) are met. The strong associations between the assessed genetic polymorphisms and the studied exposures sug- gested that the genetic instruments studied were relevant for the studied exposures. The strategy of selecting instru- ments from within the cis-region of the exposure of interest is an established approach for investigating drug effects [36]. cis-Mendelian randomisation studies are by design less prone to horizontal pleiotropy and exposure misspeci- fication (which may lead to violations of independence and exclusion restriction assumptions), as genetic variants typically exert the strongest influence on nearby genes and therefore most effects are secondary to reading of the nearby genes. However, restricting the instruments to cis- variants comes at the expense of potentially missing strong trans-variants that associate with the exposure. Further- more, our co-localisation results suggest that associations between whole-blood IL2RA and IL6R expression and the risk of type 1 diabetes are unlikely to be caused by LD with a genetic variant that primarily influences the reading of other genes in the vicinity of IL2RA or IL6R loci. Some aspects of generalisability of our results are also worth mentioning. While our study did not explicitly exclude individuals from non-European ancestries, the original GWAS studies primarily included individuals of European descent, which limits the spectrum of rare variants and the generalisability of our findings to other ancestries. Further- more, our Mendelian randomisation estimates represent the influence of small changes in IL2RA, IL6R and TYK2 expres- sion during the entire life course before the diagnosis of type 1 diabetes. Therefore, these effect sizes cannot be directly extrapolated to clinical trials in which the doses are larger and exposures shorter, and possibly outside a key sensitive window for disease development. Thus, natural history stud- ies and clinical prevention trials should pinpoint the optimal stage of pathogenesis at which to interfere with IL-2, IL-6 or TYK2 signalling to prevent type 1 diabetes. Finally, even if up to 50% of variability in genetic risk of type 1 diabetes is attributable to HLA-II locus [37], we could not analyse the interactions between SNPs reported here and the HLA genotype or genetic risk scores on risk of type 1 diabetes, or any sex-specific effects, as we did not have access to indi- vidual-level data. In conclusion, our results provide genetic evidence that IL-2, IL-6 and TYK2 signalling are associated with type 1 diabetes risk. Our findings suggest that clinical trials inves- tigating the efficacy of drugs such as tocilizumab (IL-6R antagonist that targets all IL-6 signalling modalities), olam- kicept (soluble gp130Fc that blocks IL-6 trans-signalling) and low-dose aldesleukin (IL-2 analogue) may be promising candidates for the prevention of type 1 diabetes. Supplementary Information The online version of this article (https:// doi. org/ 10. 1007/ s00125- 024- 06267-5) contains peer-reviewed but unedited supplementary material. Acknowledgements The authors wish to acknowledge CSC – IT Center for Science, Finland, for computational resources, the authors of original studies for summary statistics, T. Kinnunen (University of Eastern Finland) for helpful discussions, and two anonymous reviewers for insightful comments. During the preparation of this work the authors used ChatGPT version 4.0 to proofread and to improve the readability of the manuscript. After using this tool/service, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication. Data availability Analysis code can be accessed at https:// github. com/ jkosk ennie mi/ T1DSC REEN, which also includes instructions on how to download the original GWAS summary statistics. Funding This study was funded by Kyllikki ja Uolevi Lehikoisen säätiö, Foundation for Paediatric Research, Finnish Cultural Foundation, JDRF International (JJK); The University of Oulu & The Research Council of Finland Profi 326291 (VK); European Union’s Horizon 2020 research and innovation programme under grant agreement no. 848158 (Early- Cause) (VK); Wellcome Trust (225790/Z/22/Z) and the UK Research and Innovation Medical Research Council (MC_UU_00040/01) (VK). The study funders were not involved in the design of the study; the collection, analysis, and interpretation of data; writing the report; and did not impose any restrictions regarding the publication of the report. Authors’ relationships and activities The authors declare that there are no relationships or activities that might bias, or be perceived to bias, their work. Contribution statement JJK, VK and DG designed the study. JJK, VK and JL curated data. TEH and EKK analysed and visualised the data under supervision of JL, JJK and VK. TEH, EKK and JJK wrote the first draft of the manuscript. All authors reviewed and edited the manu- script and approved the final version to be published. All authors had access to data, and JJK, VK, JL, TEH and EKK accessed and verified the data. JJK, VK, TEH and EKK are guarantors for this article. Open Access This article is licensed under a Creative Commons Attri- bution 4.0 International License, which permits use, sharing, adapta- tion, 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 2676 Diabetologia (2024) 67:2667–2677 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/. References 1. Foster NC, Beck RW, Miller KM et al (2019) State of type 1 diabetes management and outcomes from the T1D exchange in 2016–2018. Diabetes Technol Ther 21(2):66–72. https:// doi. org/ 10. 1089/ dia. 2018. 0384 2. Dayan CM, Besser REJ, Oram RA et al (2021) Preventing type 1 diabetes in childhood. Science 373(6554):506–510. https:// doi. org/ 10. 1126/ scien ce. abi47 42 3. Herold KC, Bundy BN, Long SA et al (2019) An anti-CD3 anti- body, teplizumab, in relatives at risk for type 1 diabetes. N Engl J Med 381(7):603–613. https:// doi. org/ 10. 1056/ NEJMo a1902 226 4. King EA, Wade Davis J, Degner JF (2019) Are drug targets with genetic support twice as likely to be approved? Revised estimates of the impact of genetic support for drug mechanisms on the prob- ability of drug approval. PLoS Genet 15(12):1–20. https:// doi. org/ 10. 1371/ journ al. pgen. 10084 89 5. Giambartolomei C, Vukcevic D, Schadt EE et al (2014) Bayesian test for colocalisation between pairs of genetic association studies using summary statistics. PLoS Genet 10(5):e1004383. https:// doi. org/ 10. 1371/ JOURN AL. PGEN. 10043 83 6. Zuber V, Grinberg NF, Gill D et al (2022) Combining evidence from Mendelian randomization and colocalization: Review and comparison of approaches. Am J Hum Genet 109(5):767–782. https:// doi. org/ 10. 1016/j. ajhg. 2022. 04. 001 7. Robertson CC, Inshaw JRJ, Onengut-Gumuscu S et al (2021) Fine-mapping, trans-ancestral and genomic analyses iden- tify causal variants, cells, genes and drug targets for type 1 diabetes. Nat Genet 53(7):962–971. https:// doi. org/ 10. 1038/ s41588- 021- 00880-5 8. The ULTRA-DD Consortium, Fang H, De Wolf H et al (2019) A genetics-led approach defines the drug target landscape of 30 immune-related traits. Nat Genet 51(7):1082–1091. https:// doi. org/ 10. 1038/ s41588- 019- 0456-1 9. Chiou J, Geusz RJ, Okino ML et al (2021) Interpreting type 1 diabetes risk with genetics and single-cell epigenomics. Nature 594(7863):398–402. https:// doi. org/ 10. 1038/ s41586- 021- 03552-w 10. Ferkingstad E, Sulem P, Atlason BA et al (2021) Large-scale integra- tion of the plasma proteome with genetics and disease. Nat Genet 53(12):1712–1721. https:// doi. org/ 10. 1038/ s41588- 021- 00978-w 11. Võsa U, Claringbould A, Westra H-J et al (2021) Large-scale cis- and trans-eQTL analyses identify thousands of genetic loci and polygenic scores that regulate blood gene expression. Nat Genet 53(9):1300–1310. https:// doi. org/ 10. 1038/ s41588- 021- 00913-z 12. Rose-John S, Jenkins BJ, Garbers C, Moll JM, Scheller J (2023) Targeting IL-6 trans-signalling: past, present and future pros- pects. Nat Rev Immunol 23(10):666–681. https:// doi. org/ 10. 1038/ s41577- 023- 00856-y 13. Kamat MA, Blackshaw JA, Young R et al (2019) PhenoScanner V2: an expanded tool for searching human genotype-phenotype associations. Bioinforma Oxf Engl 35(22):4851–4853. https:// doi. org/ 10. 1093/ bioin forma tics/ btz469 14. Yazar S, Alquicira-Hernandez J, Wing K et al (2022) Single-cell eQTL mapping identifies cell type–specific genetic control of autoimmune disease. Science 376(6589):eabf3041. https:// doi. org/ 10. 1126/ scien ce. abf30 41 15. The GTEx Consortium (2020) The GTEx consortium atlas of genetic regulatory effects across human tissues. Science 369(6509):1318. https:// doi. org/ 10. 1126/ scien ce. aaz17 76 16. Hemani G, Zheng J, Elsworth B et al (2018) The MR-Base plat- form supports systematic causal inference across the human phe- nome. eLife 7:e34408. https:// doi. org/ 10. 7554/ eLife. 34408 17. Abbas AK, Trotta E, Simeonov DR, Marson A, Bluestone JA (2018) Revisiting IL-2: biology and therapeutic prospects. Sci Immunol 3(25):eaat1482. https:// doi. org/ 10. 1126/ sciim munol. aat14 82 18. Mitra S, Leonard WJ (2018) Biology of IL-2 and its thera- peutic modulation: Mechanisms and strategies. J Leukoc Biol 103(4):643–655. https:// doi. org/ 10. 1002/ JLB. 2RI07 17- 278R 19. Garg G, Tyler JR, Yang JHM et al (2012) Type 1 diabetes-asso- ciated IL2RA variation lowers IL-2 signaling and contributes to diminished CD4+CD25+ regulatory T cell function. J Immunol 188(9):4644–4653. https:// doi. org/ 10. 4049/ jimmu nol. 11002 72 20. the TEDDY Study Group, Vehik K, Lynch KF et al (2019) Pro- spective virome analyses in young children at increased genetic risk for type 1 diabetes. Nat Med 25(12):1865–1872. https:// doi. org/ 10. 1038/ s41591- 019- 0667-0 21. Lin J, Moradi E, Salenius K et  al (2023) Distinct transcrip- tomic profiles in children prior to the appearance of type 1 diabetes-linked islet autoantibodies and following enterovirus infection. Nat Commun 14(1):7630. https:// doi. org/ 10. 1038/ s41467- 023- 42763-9 22. Shouse AN, LaPorte KM, Malek TR (2024) Interleukin-2 signaling in the regulation of T cell biology in autoimmunity and cancer. Immunity 57(3):414–428. https:// doi. org/ 10. 1016/j. immuni. 2024. 02. 001 23. Bachmann MF, Wolint P, Walton S, Schwarz K, Oxenius A (2007) Differential role of IL-2R signaling for CD8 + T cell responses in acute and chronic viral infections. Eur J Immunol 37(6):1502– 1512. https:// doi. org/ 10. 1002/ eji. 20063 7023 24. International Inflammatory Bowel Disease Genetics Consortium, Huang H, Fang M et al (2017) Fine-mapping inflammatory bowel disease loci to single-variant resolution. Nature 547(7662):173– 178. https:// doi. org/ 10. 1038/ natur e22969 25. Type 1 Diabetes Genetics Consortium, Onengut-Gumuscu S, Chen W-M et al (2015) Fine mapping of type 1 diabetes suscepti- bility loci and evidence for colocalization of causal variants with lymphoid gene enhancers. Nat Genet 47(4):381–386. https:// doi. org/ 10. 1038/ ng. 3245 26. Tanaka T, Narazaki M, Kishimoto T (2014) IL-6 in inflamma- tion, immunity, and disease. Cold Spring Harb Perspect Biol 6(10):a016295. https:// doi. org/ 10. 1101/ cshpe rspect. a0162 95 27. Garbers C, Heink S, Korn T, Rose-John S (2018) Interleukin-6: designing specific therapeutics for a complex cytokine. Nat Rev Drug Discov 17(6):395–412. https:// doi. org/ 10. 1038/ nrd. 2018. 45 28. Kimura A, Kishimoto T (2010) IL-6: regulator of Treg/Th17 bal- ance. Eur J Immunol 40(7):1830–1835. https:// doi. org/ 10. 1002/ eji. 20104 0391 29. Spence A, Tang Q (2016) Restoring regulatory T cells in type 1 diabetes. Curr Diab Rep 16(11):110. https:// doi. org/ 10. 1007/ s11892- 016- 0807-6 30. Greenbaum CJ, Serti E, Lambert K et al (2021) IL-6 receptor blockade does not slow β cell loss in new-onset type 1 diabetes. JCI Insight 6(21):e150074. https:// doi. org/ 10. 1172/ jci. insig ht. 150074 31. Sosenko JM, Skyler JS, Beam CA et al (2013) Acceleration of the loss of the first-phase insulin response during the progression to type 1 diabetes in diabetes prevention trial-type 1 participants. Diabetes 62(12):4179–4183. https:// doi. org/ 10. 2337/ db13- 0656 32. Pellenz FM, Dieter C, Duarte GCK, Canani LH, de Souza BM, Crispim D (2021) The rs2304256 polymorphism in TYK2 gene is associated with protection for type 1 diabetes mellitus. Diabetes Metab J 45(6):899–908. https:// doi. org/ 10. 4093/ dmj. 2020. 0194 2677Diabetologia (2024) 67:2667–2677 33. Yuan S, Wang L, Zhang H et al (2023) Mendelian randomization and clinical trial evidence supports TYK2 inhibition as a thera- peutic target for autoimmune diseases. eBioMedicine 89:104488. https:// doi. org/ 10. 1016/j. ebiom. 2023. 104488 34. Kalvakolanu DV, Nallar SC, Kalakonda S (2016) Interferons: Cellular and Molecular Biology of Their Actions. In: Reference Module in Biomedical Sciences. Elsevier, Amsterdam, the Nether- lands, p B9780128012383961166. https:// doi. org/ 10. 1016/ B978- 0- 12- 801238- 3. 96116-6 35. Uhlén M, Fagerberg L, Hallström BM et  al (2015) Prot- eomics. Tissue-based map of the human proteome. Science 347(6220):1260419. https:// doi. org/ 10. 1126/ scien ce. 12604 19 36. Gill D, Georgakis M, Walker V et al (2021) Mendelian randomization for studying the effects of perturbing drug targets [version 2; peer review: 3 approved, 1 approved with reservations]. Wellcome Open Res 6(16):eCollection 2021. https:// doi. org/ 10. 12688/ wellc omeop enres. 16544.2 37. Pociot F, Lernmark Å (2016) Genetic risk factors for type 1 dia- betes. Lancet 387(10035):2331–2339. https:// doi. org/ 10. 1016/ S0140- 6736(16) 30582-7 Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Authors and Aliations Tea E. Heikkilä1  · Emilia K. Kaiser1  · Jake Lin2,3  · Dipender Gill4  · Jaakko J. Koskenniemi1,5  · Ville Karhunen6,7,8 * Ville Karhunen ville.karhunen@mrc-bsu.cam.ac.uk 1 Institute of Biomedicine, Research Centre for Integrative Physiology and Pharmacology, University of Turku, Turku, Finland 2 Faculty of Social Sciences, Tampere University, Tampere, Finland 3 Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, Sweden 4 Department of Epidemiology and Biostatistics, School of Public Health, Imperial College London, London, UK 5 Health Informatics Institute, Morsani College of Medicine, University of South Florida, Tampa, FL, USA 6 Research Unit of Population Health, Faculty of Medicine, University of Oulu, Oulu, Finland 7 Research Unit of Mathematical Sciences, Faculty of Science, University of Oulu, Oulu, Finland 8 MRC Biostatistics Unit, School of Clinical Medicine, University of Cambridge, Cambridge, UK