Diabetes Volume 67, April 2018 687

Molecular Pathways for Immune Recognition of Signal in Type 1 Diabetes

Deborah Kronenberg-Versteeg,1,2 Martin Eichmann,1 Mark A. Russell,3 Arnoud de Ru,4 Beate Hehn,5 Norkhairin Yusuf,1 Peter A. van Veelen,4 Sarah J. Richardson,3 Noel G. Morgan,3 Marius K. Lemberg,5 and Mark Peakman1,2

Diabetes 2018;67:687–696 | https://doi.org/10.2337/db17-0021

The region of preproinsulin (PPI) contains In type 1 diabetes, the pathological process of immune- epitopes targeted by HLA-A-restricted (HLA-A0201, A2402) mediated destruction of -producing b-cells leads to MUOOYADTRANSPLANTATION AND IMMUNOLOGY cytotoxic T cells as part of the pathogenesis of b-cell de- insulin deficiency and hyperglycemia (1). Multiple arms of struction in type 1 diabetes. We extended the discovery of the immune system are likely to contribute to this tissue- the PPI epitope to disease-associated HLA-B*1801 and damaging process, with strong indications that CD8+ cyto- HLA-B*3906 (risk) and HLA-A*1101 and HLA-B*3801 (pro- toxic T lymphocytes (CTLs) are a dominant killing pathway. tective) alleles, revealing that four of six alleles present Evidence includes data from preclinical models showing de- epitopes derived from the signal peptide region. During pendence of disease development on intact CD8/MHC class cotranslational translocation of PPI, its signal peptide is I mechanisms (2), supported by compelling findings in hu- cleaved and retained within the man studies, including the existence of high-risk polymor- (ER) membrane, implying it is processed for immune phic HLA class I (3); enrichment of effector CTLs recognition outside of the canonical proteasome-directed specificforb-cell targets in the circulation in new-onset pathway. Using in vitro translocation assays with specific disease (4,5); recapitulation of b-cell killing by patient- inhibitors and knockout in PPI-expressing target derived preproinsulin (PPI)-specificCTLsinvitro(4,6); cells, we show that PPI signal peptide CD8 T-cell dominance of islet infiltrates in patients, includ- requires (SPP). The intramem- ing the presence of CD8s bearing receptors specificfor brane protease SPP generates cytoplasm-proximal epito- b-cell autoantigens (7,8); hyperexpression of HLA class I, pes, which are transporter associated with antigen processing both at the RNA and levels, in residual insulin- (TAP), ER-luminal epitopes, which are TAP independent, containing islets (ICIs) in type 1 diabetes pancreatic tissue each presented by different HLA class I molecules and (9); and recent success in halting b-cell loss using immuno- N-terminal trimmed by ER aminopeptidase 1 for optimal therapy targeted at effector CD8 T cells (10). presentation. In vivo, TAP expression is significantly The potential for the immunological dialogue between upregulated and correlated with HLA class I hyperexpression CTLs and b-cells to be a key component of the development in insulin-containing islets of patients with type 1 diabetes. of type 1 diabetes has led several groups to focus on the Thus, PPI signal peptide epitopes are processed by SPP relevant molecular interactions that govern this interface, and loaded for HLA-guided immune recognition via path- including studies of HLA class I gene polymorphisms carry- ways that are enhanced during disease pathogenesis. ing modified risk of disease (HLA-A*0201,-A*1101,-A*2402,

1Department of Immunobiology, Faculty of Life Sciences and Medicine, King’s Received 5 January 2017 and accepted 10 January 2018. College London, London, U.K. This article contains Supplementary Data online at http://diabetes 2 ’ National Institute for Health Research, Biomedical Research Centre at Guy s and .diabetesjournals.org/lookup/suppl/doi:10.2337/db17-0021/-/DC1. St. Thomas’ Hospital Foundation Trust and King’s College London, London, U.K. D.K.-V. and M.E. contributed equally to this work. 3Institute of Biomedical and Clinical Science, University of Exeter Medical School, Exeter, U.K. D.K.-V. is currently affiliated with the Wellcome Trust – Medical Research Council 4Department of Immunohematology and Blood Transfusion, Leiden University Cambridge Stem Cell Institute, University of Cambridge, Cambridge, U.K. Medical Center, Leiden, the Netherlands © 2018 by the American Diabetes Association. Readers may use this article as 5Center for Molecular Biology of Heidelberg University (ZMBH), DKFZ-ZMBH Al- long as the work is properly cited, the use is educational and not for profit, and the liance, Heidelberg, Germany work is not altered. More information is available at http://www.diabetesjournals Corresponding author: Deborah Kronenberg-Versteeg, [email protected]. .org/content/license. 688 Preproinsulin Processing for Immune Recognition Diabetes Volume 67, April 2018

-B*1801,-B*3801,and-B*3906 [3,11,12]) and antigenic tar- RESEARCH DESIGN AND METHODS gets within b-cells recognized by CTLs. These have espe- Cell Lines and Epitope Discovery cially focused on PPI, which is considered a primary target K562 cells transfected with HLA-A*1101, HLA-B*1801, in b-cell autoimmunity because anti-insulin autoantibodies HLA-B*3801,orHLA-B*3906 and PPI cDNAs (K562-HLA- fi are frequently the rst disease manifestation in high-risk PPI cells) were generated as described previously (4). Ex- children (12). pression of HLA class I was confirmed by flow cytometry For immune recognition, processing and presentation of (anti-HLA-ABC antibody W6/32; Serotec, Oxford, U.K.). by MHC class I commonly results from degrada- was detected in supernatants by ELISA (DRG tion of by the proteasome (13), generating cytosolic International, Marburg, Germany). Subsequently, ;1010 – peptides of 8 16 amino acids (14), which are transported cells from each cell line were harvested and immunoaffinity into the endoplasmic reticulum (ER) lumen via the trans- purification of HLA-A1101, HLA-B1801, HLA-B3801, and porter associated with antigen processing (TAP) (15), part HLA-B3906, peptide extraction and nano high-performance of the MHC class I peptide loading complex (16). ER ami- liquid chromatography–mass spectrometry performed as nopeptidase (ERAP) 1 trims peptides to a suitable length described elsewhere (4,6,27,28). for MHC loading, if required (17), and peptide-MHC complexes are delivered to the cell surface for immune rec- Inhibitors, RNA Interference and T-Cell Clone Activation ognition. Proteasome-independent, cytoplasm-based nonca- K562-A24-PPI cells were transfected twice within 48 h with nonical antigen presentation pathways have also been 20 nmol/L small interfering RNAs (Applied Biosystems, described (18). Whereas these and the canonical route col- Foster City, CA) targeting B2M (s1852), ERAP1 (s28618), lectively require TAP to delivery peptides into the ER, signal ERAP2 (s34520), TAP1 (s13778, s13780), and TAP2 (s13781, peptide epitopes from secretory proteins may be MHC- s13782, s13783) using Lipofectamine RNAiMAX (Invitrogen, loaded independently of TAP (19) after Paisley, U.K.). Knockdown was assessed by RT-PCR us- cleavage and intramembrane by signal peptide ing TaqMan-specific primers and relative cDNA content peptidase (SPP) (20–22). Interestingly, peptides originating normalized to GAPDH . CD8 T-cell clones from the ER luminal side of the signal peptide can access for PPI3–11-HLA-A2402 and PPI15–24-HLA-A0201 (4,6) were the peptide loading complex directly (23–25), whereas cocultured with K562-HLA-PPI target cells at indicated epitopes close to the cytosol may require proteasomal trim- effector-to-target ratios (4 h), and response was measured ming and TAP for entry into the ER (26). However, the as degranulation (CD107a expression [29]) or macrophage extent to which signal peptides and these noncanonical inflammatory protein 1b (MIP-1b) release (R&D Systems, epitope-generation pathways fuel immune recognition of Minneapolis, MN). single antigens and play a role in physiological responses Site-Directed Mutagenesis remains unclear. To alter single or multiple amino acids (highlighted in Sup- This study was motivated by the need for a better un- plementary Table 1), PPI-containing plasmid pcDNA3/PPI derstanding of autoantigen processing for immune recog- was amplified using altered primers (PPI9P→L, CCC→CTC; nition of b-cells via these different routes, which could PPI12A→L, GCG→CTG; PPI15A→L, GCC→CTC; with 18- to provide novel insights into disease pathogenesis and high- 20-nucleotide overhang preceding and following the mu- light pathways susceptible to therapeutic manipulation. We tatedsite)andPfuTurboDNAPolymeraseAD(Agilent previously reported that the predominant epitope species Technologies, Santa Clara, CA); sequences were confirmed presented to CTLs by human cell lines cotransfected with before use. the INS gene (encoding PPI) and the HLA class I genes HLA- A*0201 and A*2402 derived from the signal peptide region In Vitro Transcription, Translation, and Translocation (4,6). These signal peptide–derived epitopes are recognized and Analysis of SPP Processing by patient CTLs and presented by b-cells bearing the rele- Plasmid pcDNA3/PPI was linearized with EcoRI and tran- vant HLA class I molecule. Here we examine whether PPI scribed in vitro with T7 RNA polymerase at 42°C using signal peptide is a more general source of epitopes by study- 500 mmol/L m7G(59)ppp(59)G CAP analog (New England ing additional HLA class I molecules associated with type 1 Biolabs, Ipswich, MA) (30). mRNAs were translated in vitro diabetes. We show that generation of epitopes from the PPI in 25 mL rabbit reticulocyte lysate (Promega, Madison, WI) signal peptide is driven by the intramembrane protease containing [35S]-methionine and [35S]-cysteine (PerkinElmer, SPP and that loading into nascent HLA class I molecules Waltham, MA) and, where indicated, two equivalents of requires trimming by ERAP1 and is either direct or follows nuclease-treated dog rough microsomes (31). cytoplasmic translocation and TAP, with the selected path- (Z-LL)2-ketone (Calbiochem SPP inhibitor, 5 mmol/L; Merck, way being determined by the HLA allele. We show that San Diego, CA) or DMSO control were added as indicated. the key factors ERAP1 and TAP are expressed in insulin- After 30 min at 30°C, microsomes were extracted with containing islets of patients studied post mortem after 500 mmol/L KOAc, solubilized in SDS sample buffer (32) type 1 diabetes diagnosis, indicating that the multiple path- and analyzed by SDS-PAGE using Tris-bicine-urea acrylam- ways that are potentially critical in the CTL–b-cell dialogue ide gels (15% Tris, 5% bicine; 8 mol/L urea) (33) and an FLA are active in the disease setting. 7000 phosphorimager (Fuji) with Multi Gauge software (Fuji). diabetes.diabetesjournals.org Kronenberg-Versteeg and Associates 689

The reference peptide comprising the 24– PPI amplification using primers (forward: 59-ATATATGAATT- signal sequence was translated in wheat germ extract (34). CGCACCCTCGCCATG-39;reverse:59-ATATATCTCGAGGC- ACCAGCTGCATCATTTC-39)(Eurofins Scientific, Ebersberg, Immunohistochemistry and Immunofluorescence Germany) and Phusion High-Fidelity DNA Polymerase (New fi fi Formalin- xed, paraf n-embedded pancreas sections from England Biolabs) and by agarose gel electrophoresis. six control and six case subjects with type 1 diabetes (Exeter Archival Diabetes Biobank, http://foulis.vub.ac.be/; RESULTS Supplementary Table 2; ethical approval 15/W/0258) were studied with the use of immunohistochemistry. Sections were Presentation of PPI Epitopes by HLA-A1101, HLA-B1801, dewaxed, rehydrated, and heated in a microwave (800 W) for HLA-B3801, and HLA-B3906 INS 20 min, then blocked in 5% normal goat serum before being K562 cells transfected with encoding human PPI and HLA-A*1101 HLA-B*1801 HLA-B*3801 HLA- incubated with primary and secondary antibodies (Supplemen- one of , , ,and B*3906 tary Table 3), counterstained with hematoxylin, dehydrated, generated surrogate b-cells secreting proinsulin and and mounted in a distyrene/xylene-based mountant (DPX). expressing relevant HLA class I molecules (Supplementary fi fi Multiple antigens within the same formalin-fixed, paraffin- Fig. 1). The immunopeptidome eluted from af nity-puri ed fi embedded section were probed sequentially with up to three HLA-A1101 identi ed 905 peptides; from HLA-B1801, 615 different antibodies (Supplementary Table 3) and images cap- peptides; from HLA-B3801, 455 peptides; and from HLA- $ tured (AF6000 system; Leica Microsystems, Milton Keynes, B3906, 298 peptides (all Mascot scores 40), and corre- U.K.). Mean fluorescence intensity of stained antigens was sponded to published HLA-binding motifs and peptide analyzed using ImageJ software, and isotype control antisera ligandomes (35). Of interest, PPI epitopes from the signal fi were used to confirm reagent specificity. peptide region were identi ed for HLA-B3801, PPI5–14 MRLLPLLALL, and HLA-B3906, PPI5–12 MRLLPLLA (Fig. SPP Knockout Using CRISPR 1 and Supplementary Fig. 2). In addition, we identified CRISPR guide sequences for exons of SPP (HM13 gene) a B-chain epitope for HLA-B3801, PPI33–41 SHLVEALYL, and were designed using CRISPR DESIGN tool (crispr.mit.edu) a C-peptide epitope for HLA-A1101, PPI80–88 LALEGSLQK. (Supplementary Table 4) and cloned into pLG2C vector Peptide identities were confirmed by tandem mass spectrom- containing EGFP linked to a Cas9 cassette via P2A. K562- etry profiling of the synthetic compound (Supplementary Fig. A2-PPI was transfected with each of two pLG2C vectors (con- 2). No other peptides from PPI were identified. taining guide sequences for exon 2 or 3 and exon 10 or 11) When considered with our previous reports of immuno- – or empty vector (mock treated, no guide sequence) using dominant PPI signal peptide derived epitopes in PPI15–24- Effectene (Qiagen, Hilden, Germany). Single cells sorted for HLA-A0201 (6) and PPI3–11-HLA-A2402 (4), these new high HLA-A2 (W6/32; BioLegend, San Diego, CA) and EGFP discoveries indicate that the signal peptide region of PPI is were examined for gene truncation by RT-PCR and for SPP a rich source for processing for immune recognition (Fig. 1

Figure 1—PPI epitope discovery in HLA-B3906. A: Tandem mass spectrometry analysis of collision-induced dissociation revealing the tandem mass spectrum of a peptide (463.8 m/z, sequence MRLLPLLA). The correct identity of the peptide was proven by tandem mass spectrometry of the synthetic compound. B: The table lists the amino acid sequences of the peptide with the expected b- and y-fragment ions (extending from the amino terminus and carboxyl terminus, respectively). Observed fragment ions are underlined. Tandem mass spectra and a table of b- and y-fragment ions for PPI epitopes discovered in HLA-A1101 and HLA-B3801 are provided in Supplementary Fig. 2. C: Diagram indicates the position of the PPI signal peptide in the ER membrane during cotranslational translocation and the sequences and positions of the signal peptides identified by elution of the naturally processed and presented immunopeptidome for different HLA molecules. 690 Preproinsulin Processing for Immune Recognition Diabetes Volume 67, April 2018

Table 1—PPI epitopes identified by elution from specific HLA class I molecules Allele Risk/protection PPI signal peptide Other PPI regions HLA-A*0201 Risk/neutral PPI15–24 (ALWGPDPAAA) No peptides found HLA-A*2402 Risk PPI3–11 (LWMRLLPLL) No peptides found HLA-B*1801 Risk No peptides found No peptides found HLA-B*3906 Risk PPI5–12 (MRLLPLLA) No peptides found HLA-A*1101 Protection No peptides found PPI80–88 (LALEGSLQK) HLA-B*3801 Protection PPI5–14 (MRLLPLLALL) PPI33–41 (SHLVEALYL) Boldface data signifies identified epitopes. and Table 1). To examine whether this arises because signal in processing and cleavage of PPI signal peptide. To test peptide regions are immunogenic per se, as has been reported this, mRNAs encoding PPI were translated in vitro with (36), the available peptidome data were mined in greater ER-derived microsomes and [35S]-labeled methionine and depth, showing that the degree to which signal peptides are cysteine. Upon insertion of PPI into microsomes, signal presented is dependent on the HLA allele. While, for example, peptidase cleaves the PPI signal sequence, liberating trans- HLA-A0201 presents a signal peptide–derived epitope from located proinsulin from its signal peptide (Fig. 2A). More- 46% of source proteins that contain a signal peptide, for over, traces of a peptide that comigrated with an in vitro HLA-A1101 this figure is only 6.9% (Supplementary Table translated reference peptide comprising the PPI signal se- 5). Probing for any signal peptide bias using in silico pre- quence were detected in the membrane fraction (Fig. 2A; diction algorithms provided similar results (Supplementary compare lanes 2 and 4). Because processing of nascent chains Table 6). We conclude that presentation of signal peptides by signal peptidase is a well-known activity in ER-derived of PPI is not likely to result from a generalized propensity of microsomes (31), and because no low–molecular weight HLA molecules to select this region for presentation, but the peptides were observed in the translation reaction lacking fi number of HLA molecules studied to date remains limited. microsomes (lane 1), we conclude that the identi ed pep- tide corresponds to traces of PPI signal peptide that re- Intramembrane Protease SPP Cleaves PPI Signal Peptide main in the ER membrane fraction. By contrast, the PPI Based on previous studies of intramembrane cleavage of signal peptide is markedly stabilized and retained in the signal peptides (37), we hypothesized that SPP is involved microsome fraction upon treatment with the SPP inhibitor

Figure 2—Processing of PPI signal peptide by SPP in a cell-free translocation assay. A: In vitro translation of wild-type (wt) PPI mRNA in the absence (lane 1) or presence of ER-derived microsomes (lanes 2 and 3) and SPP inhibitor (Z-LL)2-ketone (lane 3). Microsomes containing radiolabeled translocated proinsulin (PI) and membrane integral signal peptides (SPs) were isolated and analyzed by SDS-PAGE and audio- radiography. Lane 4 shows in vitro translated reference peptide comprising the PPI signal sequence. Upon ER targeting and signal peptidase cleavage, liberated SP is released from the membrane fraction (lane 2) in a process that can be blocked in the presence of (Z-LL)2-ketone inhibitor (lane 3), indicating SPP-catalyzed cleavage. B: In vitro translation of mutant PPI mRNA (P9L, A12L, A15L) under similar conditions shows cleavage-deficient SP retained in the membrane fraction irrespective of (Z-LL)2-ketone inhibitor. C: Wild-type (top) and P9L, A12L, A15L SP mutant (bottom) sequences. D: Quantification of PPI SP processing (mean, n =5[wt]orn = 2 [mutations]). Means indicate the relative amounts of SP obtained in lane 2 compared with the corresponding lane 3, where SPP was inhibited. Processing was quantified by comparing the intensity of SP in the absence (DMSO) and presence of the SPP inhibitor (Z-LL)2-ketone for each condition, given by the formula 100 2 (100x/y), where x is the intensity of the test band and y is the intensity of the band in the presence of full inhibition with (Z-LL)2-ketone. Equal translocation efficiency was controlled by comparing the amount of PI between conditions. See Supplementary Fig. 3 and Supplementary Table 1fordetails.A,alanine;L,;P,proline. diabetes.diabetesjournals.org Kronenberg-Versteeg and Associates 691

(Z-LL)2-ketone (Fig. 2A, lane 3). Overall, this shows that PPI presence of K562-A2-PPI-SPPko cells compared with behaves as a canonical nascent chain processed by signal mock-treated lines (Fig. 3). The difference in T-cell activation peptidase, liberating a signal peptide released from the ER is not due to differences in HLA class I expression levels membrane by SPP-catalyzed intramembrane cleavage. (Supplementary Fig. 5), and the SPP knockout phenotype of To further study the role of intramembrane proteolysis reduced T-cell activation is rescued when cells are pulsed in turnover of PPI signal peptide, amino acid residues de- with cognate peptide (Supplementary Fig. 6). These data stabilizing the helical transmembrane span surrounding the show that SPP-catalyzed processing of PPI signal peptide scissile peptide bond were mutated to leucine, which blocks is an essential step to generate an immunologically relevant SPP-catalyzed cleavage (21). Consistent with our previous epitope that is implicated as having a pathogenic role in analysis of model signal peptides, single mutation of proline type 1 diabetes through activation of b-cell-specificautor- (at position 9) and alanine (position 12 and position 15) eactive CD8 T cells (6,8). shows marginal effects on PPI signal peptide processing, Requirements for Proteasome, TAP, and ERAP in PPI whereas double mutants (9L/12L, 9L/15L, and 12L/15L) Processing and the triple mutant (9L/12L/15L) show marked inhibi- We previously showed that PPI15–24 processing and presen- tion of SPP-catalyzed processing (Fig. 2B–D, Supplementary tation by HLA-A0201 does not require proteasome cleavage fi Fig. 3, and Supplementary Table 1). This identi es PPI sig- or import into the ER via TAP (6). In contrast with PPI15–24- fi fi nal peptide as a bona de SPP substrate and con rms the HLA-A0201, however, PPI3–11-HLA-A2402 presentation is requirement for helix breaks and limited hydrophobicity for TAP dependent. RNA interference (RNAi) inhibition of TAP1 2 intramembrane cleavage. Replacement of the classic helix- (94.5 6 2.3% mRNA knockdown) and TAP2 (95.4 6 0.3% break residue proline at position 9 of PPI signal peptide mRNA knockdown) expression in K562-A24-PPI cells mark- alone is not sufficient to completely block SPP-catalyzed edly reduces MIP-1b production (60.7% and 30.2%, respec- processing and only shows effect when at least one of the tively) by the PPI3–11-HLA-A2402-specific CD8 clone upon nearby alanine residues, which show an intermediate stabil- coculture with TAP RNAi–treated K562-A24-PPI cells (Fig. ity within transmembrane helices (38), is also mutated to 4A). This is attributable to a reduction in surface density of leucine. the specific peptide HLA ligand PPI3–11-HLA-A2402 (TAP We next examined whether abrogation of SPP-catalyzed RNAi–treated K562-A24-PPI cells show only minimal reduc- cleavage of PPI signal peptide affects immune recognition of tion in HLA class I expression; Supplementary Fig. 7). signal peptide–derived epitopes. Double-targeting CRISPR- Next we investigated whether aminopeptidases residing Cas9 technology generated three independent K562-A2-PPI in the ER are involved in the processing and presentation of cell lines with SPP knockout (K562-A2-PPI-SPPko; SPP PPI signal peptide epitopes. Knockdown of ERAP1 (87.2 6 knockout validation shown in Supplementary Fig. 4). When 0.5% mRNA knockdown) in K562-A24-PPI cells by RNAi fi cultured with the PPI15–24-speci c, HLA-A0201-restricted markedly reduced MIP-1b production (62.6%) by PPI3–11- CD8 T-cell clone, degranulation (which measures clone ac- HLA-A2402-specific clone 4C6, whereas no effect was seen tivation via T-cell receptor ligation by peptide-HLA) fre- inthepresenceofERAP2 knockdown (73.7 6 6.1% mRNA quency and magnitude were markedly reduced in the knockdown; 8.0% change in MIP-1b production) (Fig. 4B).

Figure 3—Impact of SPP knockout in PPI- expressing cells upon HLA presentation of signal peptide–derived epitope presentation. The PPI15–24- specific, HLA-A0201-restricted T-cell clone was cultured at a 2:1 target-to-effector ratio, with different cell lines representing K562-A2-PPI (positive control; presents PPI15–24 in HLA-A0201); K562-A2 (negative control; lacks PPI); K562-A2-PPI-SPPmock (negative controls; two lines of mock CRISPR-Cas9 manipulation; see RESEARCH DESIGN AND METHODS); and K562-A2-PPI-SPPko (test conditions; three independent lines of CRISPR-Cas9 knockout of SPP). SPP knockout markedly reduces activation of the T-cell clone, as shown by the lower median fluorescence intensity (MFI) (B) and the reduced percentage expression (A) of CD107a. Bars and error bars represent the mean and SEM, respectively, of four independent experiments. 692 Preproinsulin Processing for Immune Recognition Diabetes Volume 67, April 2018

clone 4C6 activation is likely to be an effect on target cell fi surface density of the speci cPPI3–11-HLA-A2402 ligand.

TAP and ERAP1 Expression in Pancreas Samples From Subjects With Type 1 Diabetes Expression of the processing proteins TAP1 and ERAP1— both of which we have shown to potentially contribute to the generation of PPI epitopes that target b-cells for killing by PPI-specific cytotoxic CD8 T cells—were investigated in human pancreas recovered from healthy control subjects and those with type 1 diabetes. TAP1 was present at low levels in the pancreatic islets of healthy control subjects and was detected at similarly low levels in the insulin-deficient islets (IDIs) of patients with type 1 diabetes (Fig. 5A). By contrast, TAP1 was markedly upregulated in the ICIs of these patients (Fig. 5A). Colocalization studies revealed that the elevation in TAP1 expression was most evident in b-cells, although it was also increased in other islet cells. The immunostaining in six age-matched control subjects and six subjects with type 1 diabetes was quantified by measuring the mean fluorescence intensity of TAP1 labeling in three islets from each control subject and in six islets from each patient (three ICIs and three IDIs). This con- firmed significant elevation of TAP1 expression in ICIs from patients with type 1 diabetes compared with IDIs or control islets (P , 0.001; Fig. 5B). Application of similar approaches demonstrated that, as previously described (9), HLA-ABC was also markedly elevated in the ICIs of patients with type 1 diabetes (Fig. 5C). Moreover, a strong positive correlation existed between the expression of TAP1 and HLA-ABC in ICIs (R2 = 0.519, P , 0.001; Fig. 5D). ERAP1 was detected in the islets of healthy control subjects and in the islets of donors with type 1 diabetes (Supplementary Table 2). In contrast to TAP1, however, ERAP1 expression was not noticeably altered between the islets of control subjects and those of patients with type 1 diabetes (Supplementary Fig. 8). Figure 4—Impact of siRNA knockdown of TAP and ERAP1 on epitope presentation. Percentage mRNA knockdown in K562-A24-PPI cells Model for PPI Signal Peptide Immune Processing and treated with siRNA knockdown for b2M, TAP1,andTAP2 genes (A, left panel) and the ERAP1 and ERAP2 genes (B, left panel), and the Presentation resulting effect on MIP-1b production by the HLA-A2402-restricted, Collectively, these findings imply a model of PPI signal fi PPI3–11-speci c CD8 T-cell clone 4C6 upon coculture with the differ- peptide processing for HLA class I presentation in which A B ent cells lines ( and , right panels). Scrambled siRNAs are used as the fate that follows intramembrane cleavage depends on the control. DCt values of targeted knockdown were compared with DCt values of scrambled siRNA and expressed as the relative mRNA location within the ER membrane and HLA binding po- level of specificgeneknockdown.Efficiency of relevant mRNA knock- tential (Fig. 6). On the one hand, N-terminal peptides may TAP1 TAP2 down is high, resulting in reduced PPI3–11 presentation for , , be released into the cytoplasm, where they are dependent ERAP1 ERAP2 b2M and , but not . served as a positive control. Bars on TAP transport into the ER and ERAP1 trimming before and error bars represent the mean and SEM, respectively, of technical duplicates from three independent experiments for mRNA knockdown loading into nascent HLA molecules. As an alternative, distally and one representative example of MIP-1b production (two further generated ER luminal peptides are directly released into the repeats are shown in Supplementary Fig. 9). ER lumen, omitting the need for TAP, but these may require N-terminal trimming by ERAP1 for optimal presentation.

DISCUSSION Similar effects were seen in target killing assays (data not In this study we extend our previous PPI epitope discovery shown). RNAi for ERAP1 and ERAP2 had only minimal effort to encompass new HLA class I alleles associated with effects on total surface HLA class I expression (Supplementary risk for/protection against type 1 diabetes. This reveals PPI Fig. 7), indicating that the ERAP1-mediated interference in signal peptide as a frequent source of epitopes for multiple diabetes.diabetesjournals.org Kronenberg-Versteeg and Associates 693

HLA-A and HLA-B alleles. Our previous finding of protea- endogenous pathway of PPI presentation as it may occur in some independence for HLA-A0201 PPI presentation ahumanb-cell in vivo. Whether this approach biases epi- implied a noncanonical pathway to generate signal peptide– tope discovery toward a specific PPI region remains open to derived epitopes (6). This concept is extended in the current interpretation, although our finding of epitopes in the sig- study, in which we highlight the requirement for intramem- nal peptide, B chain, and C-peptide across the different HLA branecleavageandindicatetheimportanceofSPPinthis molecules suggests that this is unlikely. K562 cells predom- role. We further show that after intramembrane cleavage, inantly express the constitutive proteasome (39), while the pathway of peptide loading varies according to whether some evidence indicates that the immunoproteasome can asignalpeptide–derived epitope is N-terminal (requiring be upregulated in b-cells under inflammatory conditions TAP) or COOH-terminal (independent of TAP). Loading is (40), and the balance of these two could influence the spec- dictated by the HLA allele and optimized in the presence of trum of epitopes generated in human islets. Our approach ERAP1. These findings highlight a distinct set of processing is pragmatic in that obtaining sufficient, pure b-cells for principles for PPI (stoichiometry with translation and prox- such work presents a severe technical and logistical chal- imity to the site of HLA molecule synthesis and loading) lenge. We therefore elected to conduct epitope discovery that contrast with those in canonical endogenous antigen using “surrogate b-cells” and then confirm findings using processing, which relies on proteasome degradation of ef- human tissue. Our previous experience has been that epi- fete or damaged cytoplasmic proteins. Together with evi- topes identified in this way for HLA-A0201 and HLA-A2402 dence that TAP expression is upregulated in relevant tissues are faithful phenocopies of PPI presentation by b-cells. In in the disease setting, this study offers important insight both cases we were able to generate CD8 T-cell clones that into the molecular processes that are a key underpinnings recognize PPI epitopes presented by both the surrogate of interactions between cytotoxic CD8 T cells and b-cells and donor-derived b-cells. Providing similar proof for the during disease pathogenesis. less common alleles remains challenging, however (41), al- One caveat in our study relates to using tumor cells though studies examining the frequency and antigen expe- transfected with the INS gene and selected HLA alleles rience of, for example, PPI5–12-B3906 and PPI5–14-B3801 as “surrogate b-cells” to represent and understand the restricted T cells in the blood using peptide-HLA multimer

Figure 5—Expression of TAP and ERAP1 in human pancreas. Pancreas samples from six control individuals and six patients with type 1 diabetes (Supplementary Table 2) were stained for TAP1, HLA-ABC, and insulin. A: Images are representative (same microscope and camera settings) of 18 islets from control case subjects, plus 18 IDIs and 18 ICIs from patients with type 1 diabetes. The mean fluorescence intensity (MFI) for each antigen was determined within all imaged islets. B: The analysis reveals a significant increase in TAP1 expression in ICIs from individuals with type 1 diabetes (T1D) compared with IDIs or control islets. C: Consistent with previous reports, a similar change was seen with HLA-ABC expression. D: Comparing TAP1 and HLA-ABC expression within the same islets shows a positive correlation. Bars represent the mean 6 SEM; one-way ANOVA and a subsequent post hoc Tukey multiple comparison test were used to determine significance between groups (***P , 0.001). 694 Preproinsulin Processing for Immune Recognition Diabetes Volume 67, April 2018

Figure 6—Processing of the PPI signal peptide. Model of PPI signal peptide processing in which N-terminal peptides may be released by SPP into the cytoplasm, where they depend on TAP transport into the ER and ERAP1 trimming before loading into nascent HLA (illustrated by the example of PPI3–11 loading into HLA-A2402). By contrast, ER luminal peptides are released directly into the ER lumen and do not require TAP transport but equally require trimming by ERAP1 for optimal presentation (illustrated by the example of PPI15–24 loading into HLA-A0201). TCR, T-cell receptor.

technology, will provide supportive evidence of a pathogenic that is required for HLA-E-mediated immunosurveillance role. (22). This is not the first study to draw attention to the One of our findings is the identification of a noncanonical potential requirement for intramembrane cleavage of an pathway of endogenous antigen processing for an epitope autoantigen in type 1 diabetes. Using algorithms to predict relevant in the recognition of b-cells by the immune system. HLA-A0201 epitopes of islet amyloid polypeptide (IAPP), Central to this is an intramembrane cleavage step. Our studies previous studies have shown CD8 T-cell reactivity to IAPP9–17 conducted in vitro using microsomes indicate that SPP is and IAPP5–13 in the blood (43,44) and, in the case of IAPP5–13, capable of cleaving the PPI signal peptide, which is released also in islets of patients with type 1 diabetes (8). Whether from the ER membrane by SPP-catalyzed cleavage. CRISPR- these peptides are naturally processed and presented by Cas9-mediated SPP knockout indicates that this enzyme is b-cells remains unclear, however, and if confirmed for both also rate limiting for epitope generation in vivo. Using this it would imply a complex processing pathway, because the and the (Z-LL)2-ketone inhibitor approaches we are able to C-terminus of IAPP5–13 overlaps and extends into the argue that SPP knockout affects the processing of signal pep- N-terminus of IAPP9–17. tide, leading to reduced presentation of PPI15–24 and, in turn, Signal sequences are essential for to reduced activation of the specificT-cellclone.Overall,this the secretory pathway via the ER (45,46), where, after in- seems to be compelling evidence for a direct role of SPP in the sertion into the protein-conduction translocation channel, processing of PPI signal peptide in a manner that is critical to they are cleaved from the preprotein by signal peptidase the generation of this important epitope. (47); in the case of PPI15–24-HLA-A0201, this also generates SPP belongs to the family of GxGD intramembrane the epitope C-terminus. Signal peptides spanning the ER proteases including presenilin/g-secretase and related SPP- membrane require cleavage by SPP for efficient disposal like proteases (37). SPP residing in the ER cleaves various (19,20). No consensus SPP cleavage motif has been iden- signal peptides and type II membrane proteins in a number tified, beyond a strong preference for helix-destabilizing of physiological settings (42), including the generation of residues in the membrane-spanning region (21,48,49), which a regulatory peptide from the HLA-A0301 signal peptide are likely to be the proline and two alanine residues at diabetes.diabetesjournals.org Kronenberg-Versteeg and Associates 695 positions 9, 12, and 15, respectively. Indeed, in our studies, 3. Nejentsev S, Howson JM, Walker NM, et al.; Wellcome Trust Case Control mutation of any one of these has a mild effect on signal Consortium. Localization of type 1 diabetes susceptibility to the MHC class I genes peptide cleavage; any two mutations together gives HLA-B and HLA-A. Nature 2007;450:887–892 a moderate phenotype; and when all three are mutated, 4. Kronenberg D, Knight RR, Estorninho M, et al. Circulating preproinsulin signal peptide-specific CD8 T cells restricted by the susceptibility molecule HLA-A24 SPP is unable to cleave, indicating that these are the helix- are expanded at onset of type 1 diabetes and kill b-cells. Diabetes 2012;61: destabilizing residues in PPI. Once liberated, the SPP- 1752–1759 cleaved N-terminal signal peptide fragment gets access 5. Luce S, Lemonnier F, Briand JP, et al. Single insulin-specific CD8+ T cells to the cytosol, where it is either trimmed by the protea- show characteristic gene expression profiles in human type 1 diabetes. Diabetes some or directly loaded onto TAP and the MHC peptide- 2011;60:3289–3299 loading complex. In the case of PPI3–11-HLA-A2402, we 6. Skowera A, Ellis RJ, Varela-Calviño R, et al. CTLs are targeted to kill beta cells show a requirement for TAP, whereas PPI15–24-HLA-A0201 in patients with type 1 diabetes through recognition of a glucose-regulated pre- requires neither proteasome nor TAP (6), leading us to proinsulin epitope [published correction appears in J Clin Invest 2009;119:2844]. J speculate that “untapped” loading of a COOH-terminal sig- Clin Invest 2008;118:3390–3402 nal peptide fragment makes use of other chaperones, such 7. Willcox A, Richardson SJ, Bone AJ, Foulis AK, Morgan NG. Analysis of islet fl – as TAPBP. in ammation in human type 1 diabetes. Clin Exp Immunol 2009;155:173 181 8. Coppieters KT, Dotta F, Amirian N, et al. Demonstration of islet-autoreactive We considered it important to try to link findings in CD8 T cells in insulitic lesions from recent onset and long-term type 1 diabetes relation to mechanisms of b-cell antigen presentation ob- – fl patients. J Exp Med 2012;209:51 60 tained in vitro with the tissue in ammatory process taking 9. Richardson SJ, Rodriguez-Calvo T, Gerling IC, et al. Islet cell hyperexpression of place in subjects with type 1 diabetes. Our analysis shows HLA class I antigens: a defining feature in type 1 diabetes. Diabetologia 2016;59: that TAP expression is significantly increased in ICIs and 2448–2458 correlated in its hyperexpression with HLA class I molecule 10. Rigby MR, DiMeglio LA, Rendell MS, et al.; T1DAL Study Team. Targeting of expression. That the b-cell contributes to its own destruc- memory T cells with alefacept in new-onset type 1 diabetes (T1DAL study): tion in various ways has become a much-vaunted metaphor. 12 month results of a randomised, double-blind, placebo-controlled phase 2 trial. In the particular setting of PPI presentation to cytotoxic Lancet Diabetes Endocrinol 2013;1:284–294 T cells, it would seem that hyperexpression of TAP is an- 11. Howson JM, Walker NM, Clayton D, Todd JA; Type 1 Diabetes Genetics other example, certainly for HLA alleles such as HLA- Consortium. Confirmation of HLA class II independent type 1 diabetes associations in A2402. In this way it may complement the effects of the major histocompatibility complex including HLA-B and HLA-A. Diabetes Obes Metab 2009;11(Suppl. 1):31–45 hyperexpression of HLA class I and hyperglycemic condi- 12. Valdes AM, Erlich HA, Noble JA. Human leukocyte antigen class I B and C loci tions (which upregulate PPI presentation [6]) in enhancing b-cell fl contribute to type 1 diabetes (T1D) susceptibility and age at T1D onset. Hum Im- cytotoxicity under the in ammatory milieu that prevails in munol 2005;66:301–313 islets of Langerhans in patients with type 1 diabetes. 13. Townsend A, Bastin J, Gould K, et al. Defective presentation to class I-restricted cytotoxic T lymphocytes in vaccinia-infected cells is overcome by enhanced deg- radation of antigen. J Exp Med 1988;168:1211–1224 Acknowledgments. The authors are grateful to John Todd, University of 14. Chang SC, Momburg F, Bhutani N, Goldberg AL. The ER aminopeptidase, Oxford, for providing the typed HLA class I cell lines and Dr. Pierre Vantourout, King’s ERAP1, trims precursors to lengths of MHC class I peptides by a “molecular ruler” College London, for providing the pLG2C vector. mechanism. Proc Natl Acad Sci U S A 2005;102:17107–17112 Funding. This study was supported by the National Institute for Health Research 15. Spies T, Bresnahan M, Bahram S, et al. A gene in the human major histo- Biomedical Research Centre at Guy’s and St Thomas’ Hospital Trusts and King’s compatibility complex class II region controlling the class I antigen presentation College London (through a PhD studentship to D.K.-V.), JDRF (Centre Grant no. pathway. Nature 1990;348:744–747 1-2007-1803 to M.P. and a Career Development Award [5-CDA-2014-221-A-N] 16. Cresswell P. Intracellular surveillance: controlling the assembly of MHC class to S.J.R.), Diabetes UK (project grant 15/0005156 to N.G.M. and S.J.R.), and the I-peptide complexes. Traffic 2000;1:301–305 Deutsche Forschungsgemeinschaft (project grant FOR2290-TP1 to M.K.L.). 17. Saric T, Chang SC, Hattori A, et al. An IFN-gamma-induced aminopeptidase in Duality of Interest. No potential conflicts of interest relevant to this article the ER, ERAP1, trims precursors to MHC class I-presented peptides. Nat Immunol were reported. 2002;3:1169–1176 Author Contributions. D.K.-V. and M.E. designed and performed experi- 18. Neefjes J, Jongsma ML, Paul P, Bakke O. Towards a systems understanding of ments, analyzed data, conceived ideas, oversaw research, and wrote the manuscript. MHC class I and MHC class II antigen presentation. Nat Rev Immunol 2011;11:823– M.A.R., A.d.R., B.H., and N.Y. performed experiments. P.A.v.V., S.J.R., N.G.M., M.K.L., 836 and M.P. conceived ideas, oversaw research, and wrote the manuscript. M.P. is the 19. Martoglio B, Dobberstein B. Signal sequences: more than just greasy peptides. guarantor of this work and, as such, had full access to all the data in the study and Trends Cell Biol 1998;8:410–415 takes responsibility for the integrity of the data and the accuracy of the data 20. Weihofen A, Binns K, Lemberg MK, Ashman K, Martoglio B. Identification of analysis. signal peptide peptidase, a presenilin-type . Science 2002;296: Prior Presentation. Parts of this study were presented at the Immunology of 2215–2218 Diabetes 15th International Conference, San Francisco, CA, 19–23 January 2017. 21. Lemberg MK, Martoglio B. Requirements for signal peptide peptidase-catalyzed intramembrane proteolysis. Mol Cell 2002;10:735–744 References 22. Lemberg MK, Bland FA, Weihofen A, Braud VM, Martoglio B. Intramembrane 1. Atkinson MA. The pathogenesis and natural history of type 1 diabetes. Cold proteolysis of signal peptides: an essential step in the generation of HLA-E epitopes. Spring Harb Perspect Med 2012;2. pii:a007641 J Immunol 2001;167:6441–6446 2. DiLorenzo TP, Serreze DV. The good turned ugly: immunopathogenic basis for 23. Wei ML, Cresswell P. HLA-A2 moleculesinanantigen-processingmutantcell diabetogenic CD8+ T cells in NOD mice. Immunol Rev 2005;204:250–263 contain signal sequence-derived peptides. Nature 1992;356:443–446 696 Preproinsulin Processing for Immune Recognition Diabetes Volume 67, April 2018

24. Henderson RA, Michel H, Sakaguchi K, et al. HLA-A2.1-associated peptides disease-associated human leukocyte antigen-class I molecules including HLA-B*39: from a mutant cell line: a second pathway of antigen presentation. Science 1992; 06. Tissue Antigens 2014;84:378–388 255:1264–1266 36. Kovjazin R, Volovitz I, Daon Y, et al. Signal peptides and trans-membrane re- 25. El Hage F, Stroobant V, Vergnon I, et al. Preprocalcitonin signal peptide gions are broadly immunogenic and have high CD8+ epitope densities: im- generates a cytotoxic T lymphocyte-defined tumor epitope processed by a proteasome- plications for vaccine development. Mol Immunol 2011;48:1009–1018 independent pathway. Proc Natl Acad Sci U S A 2008;105:10119–10124 37. Voss M, Schröder B, Fluhrer R. Mechanism, specificity, and physiology of signal 26. Bland FA, Lemberg MK, McMichael AJ, Martoglio B, Braud VM. Requirement of peptide peptidase (SPP) and SPP-like proteases. Biochim Biophys Acta 2013;1828: the proteasome for the trimming of signal peptide-derived epitopes presented by the 2828–2839 nonclassical major histocompatibility complex class I molecule HLA-E. J Biol Chem 38. Li SC, Deber CM. A measure of helical propensity for amino acids in membrane 2003;278:33747–33752 environments. Nat Struct Biol 1994;1:368–373 27. Meiring HD, van der Heeft E, ten Hove GJ, de Jong APJM. Nanoscale LC–MS(n): 39. Anderson KS, Zeng W, Sasada T, et al. Impaired tumor antigen processing by technical design and applications to peptide and protein analysis. J Sep Sci 2002;25: immunoproteasome-expressing CD40-activated B cells and dendritic cells. Cancer 557–568 Immunol Immunother 2011;60:857–867 28. Stepniak D, Wiesner M, de Ru AH, et al. Large-scale characterization of natural 40. Lundh M, Bugliani M, Dahlby T, et al. The immunoproteasome is induced ligands explains the unique gluten-binding properties of HLA-DQ2. J Immunol 2008; by cytokines and regulates apoptosis in human islets. J Endocrinol 2017;233:369– 180:3268–3278 379 29. Knight RR, Kronenberg D, Zhao M, et al. Human b-cell killing by autoreactive 41. Baschal EE, Baker PR, Eyring KR, Siebert JC, Jasinski JM, Eisenbarth GS. The preproinsulin-specific CD8 T cells is predominantly granule-mediated with the po- HLA-B 3906 allele imparts a high risk of diabetes only on specificHLA-DR/DQ tency dependent upon T-cell receptor avidity. Diabetes 2013;62:205–213 haplotypes. Diabetologia 2011;54:1702–1709 30. Nilsson IM, von Heijne G. Determination of the distance between the oligo- 42. Avci D, Lemberg MK. Clipping or extracting: two ways to membrane protein saccharyltransferaseactivesiteandtheendoplasmic reticulum membrane. J Biol degradation. Trends Cell Biol 2015;25:611–622 Chem 1993;268:5798–5801 43. Panagiotopoulos C, Qin H, Tan R, Verchere CB. Identification of a beta-cell- 31. Martoglio B, Hauser S, Dobberstein B. Cotranslational translocation of proteins specific HLA class I restricted epitope in type 1 diabetes. Diabetes 2003;52:2647– into microsomes derived from the rough endoplasmic reticulum of mammalian cells. 2651 In Cell Biology: A Laboratory Handbook. Celis JE, Ed. San Diego, CA, Academic Press, 44. Ouyang Q, Standifer NE, Qin H, et al. Recognition of HLA class I-restricted beta- 1998, p. 265–274 cell epitopes in type 1 diabetes. Diabetes 2006;55:3068–3074 32. Weihofen A, Lemberg MK, Ploegh HL, Bogyo M, Martoglio B. Release of signal 45. Johnson AE, van Waes MA. The : a dynamic gateway at the ER peptide fragments into the cytosol requires cleavage in the transmembrane region by membrane. Annu Rev Cell Dev Biol 1999;15:799–842 aproteaseactivitythatisspecifically blocked by a novel cysteine protease inhibitor. J 46. Rapoport TA, Rolls MM, Jungnickel B. Approaching the mechanism of protein Biol Chem 2000;275:30951–30956 transport across the ER membrane. Curr Opin Cell Biol 1996;8:499–504 33. Wiltfang J, Smirnov A, Schnierstein B, et al. Improved electrophoretic separation 47. Dalbey RE, Lively MO, Bron S, van Dijl JM. The chemistry and enzymology of and immunoblotting of beta-amyloid (A beta) peptides 1-40, 1-42, and 1-43. the type I signal peptidases. Protein Sci 1997;6:1129–1138 Electrophoresis 1997;18:527–532 48. Fluhrer R, Martin L, Klier B, et al. The a-helical content of the transmembrane 34. Lemberg MK, Martoglio B. Analysis of polypeptides by sodium dodecyl sulfate- domain of the British dementia protein-2 (Bri2) determines its processing by signal polyacrylamide gel electrophoresis alongside in vitro-generated reference peptides. peptide peptidase-like 2b (SPPL2b). J Biol Chem 2012;287:5156–5163 Anal Biochem 2003;319:327–331 49. Lemberg MK, Martoglio B. On the mechanism of SPP-catalysed intramembrane 35. Eichmann M, de Ru A, van Veelen PA, Peakman M, Kronenberg-Versteeg D. proteolysis; conformational control of peptide bond hydrolysis in the plane of the Identification and characterisation of peptide binding motifs of six autoimmune membrane. FEBS Lett 2004;564:213–218