Kerry A. McLaughlin,1 Carolyn C. Richardson,1,2 Aarthi Ravishankar,1

Cristina Brigatti,3 Daniela Liberati,4 Vito Lampasona,4 Lorenzo Piemonti,3

Diana Morgan,5 Richard G. Feltbower,5 and Michael R. Christie1,2

Identification of Tetraspanin-7as a Target of Autoantibodies inType 1 DiabetesDiabetes 2016;65:1690–1698 | DOI: 10.2337/db15-1058

The presence of autoantibodies to multiple-islet autoan-tigens confers high risk for the development of type 1diabetes. Four major autoantigens are established (insulin,glutamate decarboxylase, IA2, and zinc transporter-8),but the molecular identity of a fifth, a 38-kDa membraneglycoprotein (Glima), is unknown. Glima antibodies havebeen detectable only by immunoprecipitation from ex-tracts of radiolabeled islet or neuronal cells. We sought toidentify Glima to enable efficient assay of these autoan-tibodies. Mouse brain and lung were shown to expressGlima. Membrane glycoproteins from extracts of theseorgans were enriched by detergent phase separation,lectin affinity chromatography, and SDS-PAGE. Proteinswere also immunoaffinity purified from brain extracts usingautoantibodies from the sera of patients with diabetesbefore SDS-PAGE. Eluates from gel regions equivalent to38 kDa were analyzed by liquid chromatography–tandemmass spectrometry for protein identification. Three pro-teins were detected in samples from the brain and lungextracts, and in the immunoaffinity-purified sample, butnot in the negative control. Only tetraspanin-7, a multipasstransmembrane glycoprotein with neuroendocrine ex-pression, had physical characteristics expected of Glima.Tetraspanin-7 was confirmed as an autoantigen by dem-onstrating binding to autoantibodies in type 1 diabetes.We identify tetraspanin-7 as a target of autoimmunity indiabetes, allowing its exploitation for diabetes predictionand immunotherapy.

Detection of circulating autoantibodies to pancreatic islets(1), and identification of their molecular targets (2), has

allowed the development of high-throughput autoanti-body assays for clinical diagnosis of type 1 diabetes andidentification of individuals at risk for disease. Evidencefrom both animal studies and human trials (3,4) indicatesthat type 1 diabetes may be prevented in individuals atrisk. Hence, a range of therapies to interfere with immuneresponses has proved to be effective in preventing diseasedevelopment in animal models of diabetes (5) and in slow-ing the loss of b-cell function occurring in the monthsafter disease diagnosis in humans (6–8). There is now afocus on the development of procedures to interfere spe-cifically in immune responses that cause type 1 diabetes,requiring knowledge of the major targets of the autoim-mune response. There is no single common autoimmunetarget, and individuals differ in the antigen specificity ofthe autoimmune responses that develop. Four major hu-moral autoantigens have been identified in type 1 diabetesby defining the specificity of autoantibodies in the disease:insulin (9), glutamate decarboxylase (10), IA2 (11), and zinctransporter-8 (12). Autoantibodies to a fifth major hu-moral autoantigen, a 38-kDa glycosylated membrane pro-tein (Glima), have been detected in 19–38% of patientswith type 1 diabetes, with significantly higher prevalence(up to 50%) in children (13–15). The molecular identity ofGlima has for many years proved elusive, hampering thecharacterization of autoimmunity to the protein and thedevelopment of sensitive, specific autoantibody assays.Glima is expressed in pancreatic b-cell and neuronal celllines; is hydrophobic; is heavily N-glycosylated, havingaffinity for the lectin wheat germ agglutinin; and has a

1Diabetes Research Group, Division of Diabetes & Nutritional Sciences, King’sCollege London, London, U.K.2School of Life Sciences, University of Lincoln, Lincoln, U.K.3Diabetes Research Institute, Istituto di Ricovero e Cura a Carattere Scientifico,San Raffaele Scientific Institute, Milan, Italy4Division of Genetics and Cellular Biology, Istituto di Ricovero e Cura a CarattereScientifico, San Raffaele Scientific Institute, Milan, Italy5Division of Epidemiology & Biostatistics, School of Medicine, University of Leeds,Leeds, U.K.

Corresponding author: Michael R. Christie, mchristie@lincoln.ac.uk.

Received 31 July 2015 and accepted 1 March 2016.

This article contains Supplementary Data online at http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db15-1058/-/DC1.

© 2016 by the American Diabetes Association. Readers may use this article aslong as the work is properly cited, the use is educational and not for profit, andthe work is not altered.

1690 Diabetes Volume 65, June 2016

IMMUNOLOGY

AND

TRANSPLANTATIO

N

core protein backbone of ;22 kDa (13–15). The aim ofthis study was to take advantage of these known physicalproperties to prepare Glima-enriched extracts for the iden-tification of the autoantigen by mass spectrometry.

RESEARCH DESIGN AND METHODS

PatientsSerum samples were obtained from 40 patients with type 1diabetes (12–26 years of age) within 6 months of diagnosisfrom clinics in West Yorkshire with informed consent forscreening for high-titer Glima antibodies, from 94 additionalpatients (12–63 years of age) for assay verification, andfrom 52 individuals without diabetes as negative controlsubjects. Approval for the analysis of autoantibodies in serafrom these individuals was obtained from the Yorkshire andthe Humber–Bradford Leeds Research Ethics Committee.

Screen of Sera of Patients With Type 1 Diabetes forGlima AntibodiesGlima antibodies were detected by a modification ofimmunoprecipitation assays previously described (13–15) using the neuronal mouse cell line GT1.7 as a source ofantigen. Endogenous proteins in GT1.7 cells were labeledby incubation in methionine-free DMEM medium contain-ing 4 MBq/mL 35S-methionine for 7 h at 37°C. Cells werewashed with HEPES buffer (10 mmol/L HEPES, pH 7.4,150 mmol/L NaCl, 10 mmol/L benzamidine) and stored at280°C. Frozen cell pellets were extracted in HEPES buffercontaining 2% Triton X-100 for 2 h on ice and insolublematerial removed by centrifugation at 15,000g for 15 minat 4°C. Membrane glycoproteins were isolated by incubat-ing cell extracts with wheat germ agglutinin-agarose on icefor 30 min and, after washing in HEPES buffer containing0.5 mmol/L methionine, 100 mg/L BSA, and 0.5% TritonX-100, were eluted in the same buffer containing 0.5 mol/LN-acetyl glucosamine. Aliquots (20 mL) of eluate containing3 3 105 cpm radiolabeled protein were incubated with5 mL of test sera for 18 h at 4°C, and immune complexeswere captured on 5 mL of Protein A-Sepharose. Immuno-precipitated proteins were eluted in 15 mL of SDS-PAGELoading Buffer (Novex; Life Technologies, Paisley, U.K.) withheating at 90°C for 5 min and were subjected to SDS-PAGEon 12% polyacrylamide gels. After electrophoresis, gels wereincubated in 40% v/v methanol, 2.5% v/v acetic acid, andsubsequently in Enlightning Autoradiographic Enhancer(PerkinElmer, Coventry, U.K.), each for 30 min. Gels weredried and contacted with X-ray film (BioMax MR film;Kodak, Watford, U.K.) for up to 2 weeks. After exposure,X-ray film was developed to detect radiolabeled proteinsspecifically immunoprecipitated by sera from patients withtype 1 diabetes, with bands detected in the 38,000 Mr re-gion, indicating positivity for Glima antibodies.

Tissue Expression ScreenTo identify mouse organs containing the highest levels ofGlima for use in antigen purification, competitive bindingstudies were performed using detergent extracts of organsas unlabeled competitors with 35S-methionine–labeled

Glima for binding to Glima antibodies in serum from ahigh-titer Glima antibody–positive patient. Mouse kidney,brain, heart, liver, thyroid, muscle, salivary gland, thymus,pancreas, spleen, adrenal, pituitary, and lung were dis-sected, frozen in liquid nitrogen, and stored at 280°Cbefore extraction. Tissues were homogenized in homog-enization buffer (10 mmol/L HEPES, pH 7.4, 0.25 mol/Lsucrose, 10 mmol/L benzamidine), and membrane frac-tions sedimented by centrifugation at 15,000g for 30 minat 4°C. Supernatants were removed and pellets extracted in2% Triton X-100 extraction buffer for 2 h on ice. Extractswere centrifuged at 15,000g for 30 min at 4°C, and super-natants were collected. The protein concentrations of ex-tracts were determined using the Pierce BCA Protein AssayKit (ThermoFisher Scientific, Loughborough, U.K.).

For the competition assay, wheat germ agglutinin agaroseeluates from extracts of 35S-methionine–labeled GT1.7 cellswere prepared as described above. Aliquots (20 mL) ofGT1.7 cell glycoproteins containing 3 3 105 cpm radiolabel-ed proteins were incubated for 18 h at 4°C with 5 mL ofserum from patient 029 alone, or with 10 mL of de-tergent extracts containing 100 mg of extracted proteinfrom each mouse tissue. Immune complexes were capturedon Protein A-Sepharose and processed for SDS-PAGE andautoradiography.

Partial Purification of Glima From Mouse Brainand LungMouse brain and lung, shown to express Glima in the tissuescreen (see RESULTS), were homogenized in ice-cold homog-enization buffer in a Dounce homogenizer, and cell debriswas removed by centrifugation at 500g for 5 min at 4°C. Amembrane fraction was prepared by centrifugation of thesupernatant at 10,000g for 15 min at 4°C, and the pelletwas washed and extracted in 2% Triton X-114 extractionbuffer for 2 h at 4°C. Insoluble material was removed bycentrifugation at 10,000g for 15 min at 4°C, and a detergentphase was prepared by heat-induced phase separation aspreviously described (13). Fractions were added to wheatgerm agglutinin-agarose at a ratio of 100 mL lectin-agaroseto 5 mg total protein and incubated overnight at 4°C withgentle mixing. The beads were washed twice with HEPESbuffer containing 0.5% Triton X-100 and twice in NOGbuffer (1% N-octyl-glucopyroside in HEPES buffer). Wheatgerm agglutinin–binding proteins were eluted in 0.5 mol/LN-acetyl-glucosamine in NOG buffer.

Eluates were concentrated using a Pierce SDS-PAGESample Preparation Kit (Life Technologies), solubilized inSDS-PAGE Loading Buffer (Novex; Life Technologies) for10 min at 60°C, and electrophoresed on 12% Bis-Tris gelsin MOPS running buffer. Gels were stained with BrilliantBlue G-Colloidal Coomassie (Sigma-Aldrich, Poole, U.K.),and gel slices corresponding to the 38,000 Mr region wereexcised for mass spectrometry.

Immunoaffinity Purification With Glima AntibodiesFor immunoaffinity purification, 250 mL of pooled serafrom three patients with high levels of Glima antibodies

diabetes.diabetesjournals.org McLaughlin and Associates 1691

was used with 250 mL of sera from three antibody-negativeindividuals as a negative control. Sera were incubated withProtein A-Sepharose (250 mL) for 1 h at room temperaturewith rolling and washed three times in borate buffer (100mmol/L boric acid, pH 8.3). Antibodies were cross-linkedto Protein A-Sepharose with 20 mmol/L dimethyl pimelimi-date in borate buffer for 1 h. Unreacted sites were blockedwith 20 mmol/L ethanolamine for 10 min and washed be-fore use.

Triton X-114 detergent phase–purified amphiphilic pro-teins from mouse brains prepared as above were added tothe Glima antibody–positive and Glima antibody–negativebeads and incubated overnight at 4°C with mixing. Beadswere washed with 0.5% Triton X-100 in HEPES buffer priorto elution in 2% SDS at 90°C for 10 min. The eluate wasconcentrated to 20 mL using the SDS-PAGE SamplePreparation Kit and was subjected to SDS-PAGE andColloidal Coomassie gel staining as above, and gel slicesin the 38,000Mr region were excised for mass spectrometry.

In-Gel Trypsin Digestion and Mass Spectrometryof 38,000 Mr ProteinsGel slices representing 38,000 Mr regions of all sampleswere processed using the Pierce In-Gel Tryptic DigestionKit (ThermoFisher Scientific) according to the manufac-turer instructions, and the trypsin-treated extracts were vac-uum dried and stored at220°C prior to mass spectrometry.Samples were reconstituted in 30 mL of 50 mmol/L am-monium bicarbonate for 30 min at room temperature andcentrifuged at 15,000g for 15 min to remove insolublematerial. Samples were transferred to autosampler tubes,and 10 mL of each was analyzed by liquid chromatography–tandem mass spectrometry (LC-MS/MS). Peptides wereresolved by reversed-phase chromatography on a 75-mmC18 EASY column using a linear gradient of acetonitrile in0.1% formic acid at a flow rate of 300 nL/min over 50 minon an EASY Nano LC system (ThermoFisher Scientific). Theeluate was ionized by electrospray ionization using anOrbitrap Velos Pro mass spectrometer (ThermoFisher Scien-tific) operating under Xcalibur (version 2.2; ThermoFisherScientific) and by precursor ions selected based on theirintensity for sequencing by collision-induced fragmenta-tion. The tandem mass spectrometry (MS/MS) analyseswere conducted using collision energy profiles that werechosen based on the charge/mass ratio and the chargestate of the peptide.

Tandem mass spectra were processed into peak listsusing Proteome Discoverer (version 1.3; ThermoFisherScientific). All MS/MS samples were analyzed usingMascot (version 2.2.06; Matrix Science, London, U.K.)searching the UniProt Mus musculus database, assumingdigestion with trypsin. Mascot was searched with a frag-ment ion mass tolerance of 0.80 Da and a parent ion tol-erance of 10.0 ppm, with oxidation of methionine andcarbamidomethylation of cysteine as variable modifica-tions. Each data set was analyzed with a reverse FASTAdatabase acting as a decoy.

Scaffold (version 4.3.2; Proteome Software Inc., Port-land, OR) was used to validate MS/MS-based peptide andprotein identifications, which were assigned by PeptideProphet algorithms (16,17) and accepted at .95.0% prob-ability. The UniProt database was manually searched forthe physical characteristics of the proteins identified, in-cluding molecular weight, tissue distribution, and glycosyl-ation. Of those identified, only one, tetraspanin-7 (Tspan7),matched the known properties of Glima (see RESULTS) andwas characterized further.

ImmunohistochemistryTspan7 localization in rodent tissues was performed byimmunohistochemistry. Sections of formalin-fixed, paraffin-embedded rat brain, pituitary gland, pancreas, adrenalgland, lung, muscle, heart, liver, kidney, spleen, and thymuswere dewaxed, and subjected to epitope retrieval ina microwave pressure cooker in 10 mmol/L citric acid(pH 6.0) and 0.05% Tween 20. Endogenous peroxidaseactivity was inhibited with 0.3% H2O2, and nonspecificbinding was blocked with 25% nonimmune swine serumin PBS. Primary antibody to Tspan7 (anti-TM4SF2, catalog#HPA003140; Sigma-Aldrich) was applied at 1:1,000 di-lution and incubated overnight at 4°C. Antibody labelingwas detected with the Envision Kit (Dako, Ely, U.K.)according to the manufacturer instructions, and sectionswere counterstained in Mayer’s Hematoxylin Solution(Sigma-Aldrich) and visualized by microscopy.

Cloning and Expression of Recombinant Tspan7cDNA for the coding region of mouse Tspan7 was am-plified by RT-PCR from the mouse islet cell line Min6using primers (59-GAATTCATGGCATCGAGGAGAATGG-39 and 59-AGATCTCACCATCTCATACTGATTGGC-39) thatintroduce EcoR1 and BglII sites at the 59 and 39 ends,respectively, with the native stop codon removed to allowexpression as a fusion protein with a COOH-terminal pu-rification tag. The PCR product was cloned into the pFLAG-CTS expression vector for protein expression in Escherichiacoli BL21 cells after induction with isopropyl b-D-1-thiogalactopyranoside. Expressed protein was extracted fromcells with Hen Egg Lysozyme (1 mg/mL) in PBS containing10 mmol/L benzamidine, 1 mmol/L phenylmethylsulfonylfluoride for 30 min at room temperature, followed byincubation in Triton X-100 (0.1%) for 5 min and DNase(1 mg/mL) for 10 min. The lysate was centrifuged at10,000g for 10 min at 4°C, and the supernatant wasused in immunoprecipitation assays.

Tspan7 Binding to Autoantibodies in Type 1 DiabetesIndividual Glima antibody–positive and Glima antibody–negative human sera (15 mL) were incubated with ProteinA-Sepharose (15 mL), and the Ig captured was cross-linkedto beads with dimethyl pimelimidate (18). Bead-boundantibodies were incubated overnight at 4°C with TritonX-100 extracts of mouse brain or with lysates of E. coliexpressing recombinant mouse Tspan7. Beads were washedthree times in 0.5% Triton X-100 in HEPES buffer, and

1692 Tetraspanin-7 in Type 1 Diabetes Diabetes Volume 65, June 2016

captured proteins were subjected to SDS-PAGE andWestern blotting using rabbit anti-Tspan7 antibody(anti-TM4SF, catalog #HPA003140; Sigma-Aldrich) at1:250 dilution overnight at 4°C. ImmunoprecipitatedTspan7 was detected with goat anti-rabbit IgG-peroxidase(catalog #A0545; Sigma-Aldrich) and SuperSignal West PicoChemiluminescent substrate (ThermoFisher Scientific).

Luminescent Immunoprecipitation Assay for Detectionof Tspan7 AntibodiesThe coding region of human Tspan7 was cloned into thepCMVTnT vector as a fusion with nanoluciferase (NanoLuc)(Promega, Southampton, U.K.) at the 39 end. The constructwas transfected into HEK 293 cells with ExpiFectamine(ThermoFisher Scientific) for transient expression of an-tigen. Transfected cells were extracted in either 2%Triton X-114 or passive lysis buffer (Promega), and insol-uble material was removed by centrifugation. TritonX-114 extracts were subjected to phase separation as de-scribed above and analyzed for fusion protein by Westernblotting with rabbit antibodies to NanoLuc (a gift fromPromega) or Tspan7 (anti-TM4SF). Luciferase expressionin the cell extracts was quantified by luminometry usingNano-Glo assay reagent (Promega), and aliquots of pas-sive lysis buffer extracts containing 106 light units ofantigen were incubated with 5 mL of serum at 4°C for16 h prior to the capture of antibody complexes on Pro-tein A-Sepharose. Mouse brain extracts (150 mg protein)or lysates of E. coli expressing Tspan7 (250 mg protein)were added to reactions as sources of Tspan7 to com-pete for antibody binding. Complexes were washed andluciferase activity immunoprecipitation determined byluminometry with Nano-Glo assay reagent.

RESULTS

Selection of Glima Antibody–Positive SeraTo identify patients with high levels of Glima antibodiesfor immunoaffinity purification, sera from 40 patients withrecent-onset type 1 diabetes were screened by immuno-precipitation using radiolabeled mouse GT1.7 cell extracts(Fig. 1). Intense diffuse 38,000 Mr bands, which are indic-ative of high levels of Glima antibodies, were detected forthree patients (Fig. 1, patients 029, 037, and 110). Thesethree sera were used for subsequent Glima characterizationand purification. Weaker 38,000 Mr bands indicative ofGlima antibody positivity were detected in an additional11 patients (Fig. 1).

Tissue Specificity of Glima ExpressionTo identify large organs in which Glima is expressed atsuitably high levels for antigen purification, competi-tive binding studies were performed in which detergentextracts of normal mouse tissues acted as unlabeledcompetitors with radiolabeled Glima from GT1.7 celllysates for binding to antibodies in serum from patient029, who was strongly Glima antibody positive. Ex-tracts of brain, pituitary, and lung reduced the in-tensity of radiolabeled 38,000 Mr protein, which is

indicative of Glima immunoreactivity in these tissues(Fig. 2).

Identification of Glima Candidate Proteins byMass SpectrometryExtracts enriched for glycosylated membrane proteins fromboth brain and lung were prepared using Triton X-114phase separation of amphiphilic membrane proteins fol-lowed by wheat germ agglutinin affinity purification.Proteins migrating at 38,000 Mr by SDS-PAGE were tryp-sinized and analyzed by LC-MS/MS. A total of 65 candi-dates in brain and 25 in lung were identified, of which 20were common to both samples (Supplementary Table 1).Glycosylated membrane proteins immunoprecipitated frombrain extracts by antibodies in the high Glima antibody titerpatients’ serum pool were also subjected to SDS-PAGE andLC-MS/MS analysis, and 3 of the 20 protein candidatescommon to brain and lung were present in the Glimaantibody–positive sample, but not in the negative con-trol (Supplementary Table 1). These were as follows: 1)cytoplasmic actin-1 (Actb), a ubiquitous nonglycosylatedcytoskeletal protein with a predicted molecular weight of42 kDa; 2) guanine nucleotide-binding protein G(i) sub-unit a-2 (Gnai2), a nonglycosylated membrane–associated40-kDa protein with a wide tissue distribution; and 3)Tspan7, a hydrophobic four-transmembrane domain pro-tein with a core molecular weight of 27.5 kDa, five putativeN-glycosylation sites, and a neuroendocrine distribution.Tspan7 closely matched the known properties of Glima,and additional studies were performed to compare proper-ties and validate Tspan7 as the autoantigen.

Localization of Tspan7 in Rat TissuesThe tissue distribution of Tspan7 was determined byimmunohistochemistry for comparison with patterns of

Figure 1—Autoradiogram showing a screen of serum samples frompatients with type 1 diabetes for Glima antibodies by immunopre-cipitation of the 38,000 Mr protein from extracts of GT1.7 cells withdetection by SDS-PAGE and autoradiography. A normal controlserum (-ve) and a patient with type 1 diabetes previously deter-mined to be positive for Glima antibodies (SL) were included inthe assay. The location of Glima on the autoradiograph is marked.An indication of whether the samples were determined to be neg-ative (2) or positive (+) for Glima antibodies is shown under eachlane of the figure.

diabetes.diabetesjournals.org McLaughlin and Associates 1693

Glima expression in the competition experiments describedabove. Strong immunolabeling for Tspan7 was detected inthe rat brain, in particular the cerebral cortex, hippocam-pus, cerebellum, striatum, and thalamus (Fig. 3A); in thepancreatic islets (Fig. 3B); in the anterior pituitary (Fig.3C); and in epithelial cells lining the alveoli in the lung(Fig. 3D). These observations agree with the Glima immuno-reactivity described above and previously (13,15). WeakTspan7 immunolabeling was also found in cells of the ad-renal gland (Fig. 3E). No evidence of Tspan7 expression wasfound in the exocrine pancreas (Fig. 3B), muscle, heart, liver,kidney, spleen, or thymus (data not shown).

Immunoprecipitation of Tspan7 by Antibodiesin Type 1 DiabetesTo demonstrate that Tspan7 is a target for autoantibodiesin type 1 diabetes, extracts of mouse brain and lysates ofE. coli expressing recombinant mouse Tspan7 were subjectto immunoprecipitation with Glima antibody–positive andGlima antibody–negative sera followed by Western blot-ting with a rabbit antibody to Tspan7. A 38,000 Mr bandrepresenting Tspan7 was selectively immunoprecipi-tated from mouse brain detergent extract by three Glimaantibody–positive sera, but not by control samples. Thebands detected comigrated with the brain lysate control(Fig. 4A). Antibodies in the sera of patients with type 1diabetes also specifically immunoprecipitated Tspan7 frombacterial lysates containing recombinant protein (Fig. 4B).Here, the protein migrated at ;22,000 Mr, which is con-sistent with a lack of glycosylation in bacteria (Fig. 4B). Theresults confirm Tspan7 as a target of autoantibodies intype 1 diabetes.

Analysis of Tspan7 Antibodies by LuminescenceImmunoprecipitation AssayPatients screened for Glima antibodies were analyzed forTspan antibodies by immunoprecipitation of recombinantNanoLuc-tagged human Tspan7. Western blotting with

rabbit polyclonal antibodies to both NanoLuc and Tspan7detected diffuse 38,000Mr bands (the expected size of thenonglycosylated fusion protein) as the dominant immu-noreactivity in cells transfected with the construct, withadditional bands at approximately 80,000 Mr (Fig. 4C).The 38,000 Mr protein partitioned into the detergent ontemperature-induced phase separation in Triton X-114.Transfected cell extracts were used in immunoprecipitationstudies with normal control sera or with sera from Glimaantibody–positive and Glima antibody–negative patientswith type 1 diabetes. All but one of the control subjects(control sample V015) (n = 52) had low levels of Tspan7antibodies (Fig. 4D). Four patients with high levels ofGlima antibodies (Fig. 1) also immunoprecipitated highluciferase activity in the Tspan7 antibody assay (Fig. 4D),and significantly higher levels of Tspan7 antibodies werefound in Glima antibody–positive patients than Glimaantibody–negative patients (P , 0.0001; Mann-WhitneyU test). In competition assays, natural or recombinantTspan7 in brain or E. coli extracts partially (control sampleV015) or completely (Glima antibody–positive patientswith type 1 diabetes) blocked antibody binding to theNanoLuc-Tspan7 construct (Fig. 4E). Control sample V015did not bind Tspan7 from mouse brain extracts whentested in the Western blotting assay. A second set of 94patients with recent onset of type 1 diabetes was alsotested in the Tspan7 antibody assay. Using a cutoff ofmean 63 SDs of control subjects (omitting the outlier),40 (43%) were positive for Tspan7 antibodies (Fig. 4D).

DISCUSSION

Autoantibodies to Glima in type 1 diabetes were firstreported in 1996 (13), but its molecular identity has sincethen remained unknown. We used mass spectrometry ofGlima-enriched fractions of brain and lung to search forlikely candidates for Glima. LC-MS/MS analysis identified65 proteins in 38,000Mr gel samples of amphiphilic mem-brane glycoproteins from brain and 25 proteins fromlung, of which 20 were common to both (SupplementaryTable 1). Additional LC-MS/MS analysis of immunoaffinity-purified proteins isolated from brain extracts using ProteinA-Sepharose–coupled Igs from Glima antibody–positivesera (with similar preparations from Glima antibody–negative sera as a negative control) further narroweddown the potential candidates for the testing of auto-antigenicity. Only six proteins detected in the brain orlung extracts were also present in the immunoaffinity-purified sample but absent in the negative control. Ofthese, five (Actb, Gnai2, Sfxn5 [sideroflexin-5], Kctd12[potassium channel tetramerization domain 12], andTuba1b [tubulin a-1B chain]) had a considerably higherpredicted molecular weight (.36 kDa) than expected forthe nonglycosylated Glima protein (;22 kDa) (15). Fur-thermore, Actb, Gnai2, and Tuba1a are ubiquitous cyto-plasmic proteins lacking the amphiphilic characteristicsexpected of Glima. Sfxn5 and Kctd12 were not detectedin the lung extract. Tspan7 was, consequently, the most

Figure 2—Autoradiograph of tissue expression screen demonstrat-ing competition for Glima antibody binding to Glima antibody–positive (Glima Ab +ve) serum sample from patient 029 by proteinsin Triton X-100 detergent (100 mg) extracts of normal mouse tis-sues. Serum from an individual without diabetes was included as anegative control (-ve control). Reduced intensity of the 38,000 Mr

band is indicative of Glima immunoreactivity in that tissue.

1694 Tetraspanin-7 in Type 1 Diabetes Diabetes Volume 65, June 2016

promising candidate for Glima identified in the LS-MS/MSanalyses.

Tspan7 is a member of the tetraspanin family, membersof which share structural characteristics of four trans-membrane domains, with one short (EC1) and one long(EC2) extracellular loop (19). Four of the putative N-glycosylation sites are contained within the EC2 domainwith a further site located in EC1. The presence of multipletransmembrane domains and N-glycosylation sites withinTspan7 is consistent with the hydrophobic properties andheavy N-glycosylation previously reported for Glima (13,15).There are four amino acid differences between mouseand human Tspan7, all of which are located in the longextracellular loop. The tissue distribution of Tspan7 ex-pression has not been widely investigated, but analysisof the transcriptional activity of the Tspan7 gene hasshown restricted tissue distribution with high levels

being detected in regions of the adult mouse brain andlung (20). In the pancreas, Tspan7 is found specifically inthe islets of Langerhans (21). Functionally, tetraspaninfamily members are involved in mediating signal trans-duction events and have been noted to regulate cell de-velopment, activation, growth, and motility through thetrafficking of other transmembrane proteins (22). Boththe extracellular and intracellular domains are able to in-teract with other proteins, and a number of tetraspaninsbind integrins, thereby forming links with the actin cyto-skeleton (23). Mutations in the Tspan7 gene are associatedwith X-linked mental retardation and neuropsychiatric dis-eases, potentially as a result of impaired ability of the actincytoskeleton to drive neurite outgrowth (24). The ability oftetraspanins to form complexes with other membrane andcytosolic proteins may explain the copurification of multipleprotein fragments identified in the LC-MS/MS analysis.

Figure 3—Immunohistochemical analysis of Tspan7 expression in rat tissues. Sections of formalin-fixed, paraffin-embedded rat tissueswere labeled with rabbit anti-serum to Tspan7 and labeling detected by a peroxidase/3,3-diaminobenzidine–based system, with positivelabeling detected as brown staining under the microscope. Representative images of the labeling of tissue sections of brain (A), pancreas(B), pituitary (C), lung (D), and adrenal gland (E) are shown.

diabetes.diabetesjournals.org McLaughlin and Associates 1695

Proteins immunoprecipitated by Glima antibody–positivesera from brain extracts, or from bacterial extracts con-taining recombinant Tspan7, also bound rabbit antibodiesto Tspan7 by Western blotting, confirming Tspan7 as thetarget of the antibodies. Glima autoantibodies bound boththe glycosylated natural 38,000Mr Tspan7 in brain and the22,000 Mr nonglycosylated form of the protein expressed

in E. coli. A luminescence-based immunoprecipitation system(LIPS) (25) using NanoLuc-tagged human Tspan7 expressedin mammalian cells showed that patients with high levelsof Glima autoantibodies were also strongly positive in theanti-Tspan7 LIPS. The relationship between Glima andTspan autoreactivity is imperfect, which may in part bethe consequence of difficulties in ascertaining whether or

Figure 4—A: Tspan7 labeling of Western blots of mouse brain proteins immunoprecipitated by antibodies in sera from Glima antibody–negative (-ve) and Glima antibody–positive (+ve) patients with recent onset of type 1 diabetes. The migration of molecular weight markers(1023 3 Mr) on the gel and the localization of Tspan7-specific bands at 38,000 Mr are marked. B: Tspan7 labeling of Western blots ofproteins from lysates of Tspan7 expressing E. coli immunoprecipitated by antibodies in sera from Glima antibody–negative and Glimaantibody–positive patients with recent onset of type 1 diabetes. The localization of Tspan7-specific bands at ;22,000 Mr is marked; themigration of molecular weight markers is as in A. In both panels A and B, IgG heavy chains (50,000Mr), light chains (25,000 Mr), and cross-linked Ig (>100,000 Mr) from all serum samples were also detected on the blot as a consequence of cross-reactivity with the peroxidase-conjugated anti-rabbit detection antibody. C: Tspan7 was expressed as a fusion protein with NanoLuc, and Triton X-114 extracts of cellswere subject to heat-induced phase separation. Detergent and aqueous phases were subject to SDS-PAGE and Western blotting withantibodies to NanoLuc or Tspan7. The migration of molecular weight markers are shown (1023 3 Mr). D: Detergent extracts of NanoLuc-tagged Tspan7 were immunoprecipitated with normal control sera (Controls) (n = 30), sera from Glima antibody (Ab)–positive patients withtype 1 diabetes (T1D) (n = 15), and the sera of Glima antibody–negative patients with type 1 diabetes and luciferase activity associated witheach immunoprecipitate determined by luminometry. Data are plotted as luciferase activity immunoprecipitated in kilo light units (kLU), andsample codes for control or individuals with diabetes with high levels of antibodies are shown. E: Samples from control individuals or Glimaantibody–positive patients with type 1 diabetes were tested for competitive binding by natural or recombinant Tspan7 in the LIPS byperforming the immunoprecipitations in the absence (black bars) or presence of 150 mg of mouse brain extract (white bars) or 250 mg oflysates of E. coli expressing recombinant Tspan7 (hatched bars). Assays were performed in triplicate. The addition of brain and E. coli lysatesignificantly blocked antibody binding for all samples (P < 0.0001; ANOVA with Dunnett correction for multiple comparisons), with theexception of control sample CH.

1696 Tetraspanin-7 in Type 1 Diabetes Diabetes Volume 65, June 2016

not diffuse Glima bands are present on autoradiographs ofimmunoprecipitation reactions (Fig. 1). Individuals withoutdiabetes had low levels of Tspan7 antibodies, with theexception of one strongly positive control. Antibodies inthis control serum bound poorly to natural Tspan7 frommouse brain extracts, suggesting that antibodies in thisparticular sample may bind epitopes not displayed on thenatural protein. The SDS-PAGE gel migration of the fu-sion protein indicated that the majority of recombinantluciferase-tagged proteins were not subject to the heavyglycosylation found on the natural protein, which is in-dicative of incorrect membrane insertion, protein folding,or intracellular targeting of the fusion protein requiredfor appropriate posttranslational modification. Incorrectfolding or lack of glycosylation may reveal antibody epi-topes not normally displayed on the natural Tspan7. Fur-ther optimization of Tspan7 expression should permit thedevelopment of high-throughput assays for the detectionof diabetes-associated Tspan7 autoantibodies with highsensitivity and specificity.

Autoimmunity to major autoantigens in type 1 diabetesappears within the first 5 years of life in at-risk children(26), with individual immune responses developing se-quentially rather than simultaneously (27). Autoimmunityin the disease is therefore progressive, with the order ofappearance of autoimmune responses to individual anti-gens differing between individuals and diversification ofthe immune response being essential for disease progres-sion; therefore, disease rarely develops in individuals inwhom autoimmunity develops only to single autoantigens(28). The optimum strategy currently adopted for assessingdisease risk is to screen individuals for the presence of auto-antibodies to multiple islet autoantigens. The inclusion ofTspan7 antibodies in the screen may improve the sensitivityand specificity of disease prediction in large populations, andwill provide a fuller description of the major autoimmuneresponses that are developing in that individual, which isnecessary for guiding the selection of autoantigen-specificimmunotherapeutic agents to prevent the disease.

Acknowledgments. The authors thank Raymond Chung and MalcolmWard of King’s College London Proteomics Facility for LC-MS/MS analyses.Funding. This study was funded by a research grant from Diabetes UK (grant11/0004297) and by a Society for Endocrinology Early Career Award to K.A.M.C.C.R. was supported by a PhD Studentship from King’s College London Grad-uate School. Research by C.B., D.L., V.L., and L.P. was conducted within theframework of the Italian Ministry of Research project “Ivascomar project, Clus-ter Tecnologico Nazionale Scienze della Vita ALISEI.”Duality of Interest. No potential conflicts of interest relevant to this articlewere reported.Author Contributions. K.A.M. designed the study, researched and analyzeddata, and wrote the manuscript. C.C.R., A.R., C.B., D.L., V.L., L.P., D.M., and R.G.F.researched and analyzed data. M.R.C. designed the study, researched and analyzeddata, and contributed to the writing of the manuscript. All authors reviewed and editedthe manuscript and approved the final version for submission. M.R.C. is the guarantorof this work and, as such, had full access to all the data in the study and takesresponsibility for the integrity of the data and the accuracy of the data analysis.

Prior Presentation. Parts of this study were presented in abstract form atthe 75th Scientific Sessions of the American Diabetes Association, Boston, MA,5–9 June 2015.

References1. Bottazzo GF, Doniach D. Islet-cell antibodies (ICA) in diabetes mellitus(evidence of an autoantigen common to all cells in the islet of Langerhans). RicClin Lab 1978;8:29–382. Weenink SM, Christie MR. Autoantibodies in diabetes. In Autoantibodies andAutoimmunity: Molecular Mechanisms in Health and Disease. Pollard M, Ed.Weinheim, Germany, VCH Verlag GmbH & Co, 2005, p. 321–3493. Skyler JS. Primary and secondary prevention of type 1 diabetes. Diabet Med2013;30:161–1694. Michels A, Zhang L, Khadra A, Kushner JA, Redondo MJ, Pietropaolo M.Prediction and prevention of type 1 diabetes: update on success of prediction andstruggles at prevention. Pediatr Diabetes 2015;16:465–4845. Luo X, Herold KC, Miller SD. Immunotherapy of type 1 diabetes: where arewe and where should we be going? Immunity 2010;32:488–4996. Keymeulen B, Vandemeulebroucke E, Ziegler AG, et al. Insulin needs afterCD3-antibody therapy in new-onset type 1 diabetes. N Engl J Med 2005;352:2598–26087. Pescovitz MD, Greenbaum CJ, Krause-Steinrauf H, et al.; Type 1 DiabetesTrialNet Anti-CD20 Study Group. Rituximab, B-lymphocyte depletion, and pre-servation of beta-cell function. N Engl J Med 2009;361:2143–21528. Feutren G, Papoz L, Assan R, et al. Cyclosporin increases the rate andlength of remissions in insulin-dependent diabetes of recent onset. Results of amulticentre double-blind trial. Lancet 1986;2:119–1249. Palmer JP, Asplin CM, Clemons P, et al. Insulin antibodies in insulin-dependent diabetics before insulin treatment. Science 1983;222:1337–133910. Baekkeskov S, Aanstoot HJ, Christgau S, et al. Identification of the 64Kautoantigen in insulin-dependent diabetes as the GABA-synthesizing enzymeglutamic acid decarboxylase. Nature 1990;347:151–15611. Payton MA, Hawkes CJ, Christie MR. Relationship of the 37,000- and40,000-M(r) tryptic fragments of islet antigens in insulin-dependent diabetes tothe protein tyrosine phosphatase-like molecule IA-2 (ICA512). J Clin Invest 1995;96:1506–151112. Wenzlau JM, Juhl K, Yu L, et al. The cation efflux transporter ZnT8(Slc30A8) is a major autoantigen in human type 1 diabetes. Proc Natl Acad SciU S A 2007;104:17040–1704513. Aanstoot HJ, Kang SM, Kim J, et al. Identification and characterization ofglima 38, a glycosylated islet cell membrane antigen, which together with GAD65and IA2 marks the early phases of autoimmune response in type 1 diabetes.J Clin Invest 1996;97:2772–278314. Winnock F, Christie MR, Batstra MR, et al.; Belgian Diabetes Registry.Autoantibodies to a 38-kDa glycosylated islet cell membrane-associated antigenin (pre)type 1 diabetes: association with IA-2 and islet cell autoantibodies. Di-abetes Care 2001;24:1181–118615. Roll U, Turck CW, Gitelman SE, et al. Peptide mapping and characterisationof glycation patterns of the glima 38 antigen recognised by autoantibodies in typeI diabetic patients. Diabetologia 2000;43:598–60816. Keller A, Nesvizhskii AI, Kolker E, Aebersold R. Empirical statistical model toestimate the accuracy of peptide identifications made by MS/MS and databasesearch. Anal Chem 2002;74:5383–539217. Nesvizhskii AI, Keller A, Kolker E, Aebersold R. A statistical model for iden-tifying proteins by tandem mass spectrometry. Anal Chem 2003;75:4646–465818. Harlow E, Lane D. Antibodies: A Laboratory Manual. Cold Spring Harbor, NY,Cold Spring Harbor Laboratory Press, 198819. Hemler ME. Tetraspanin functions and associated microdomains. Nat RevMol Cell Biol 2005;6:801–81120. Lizio M, Harshbarger J, Shimoji H, et al.; FANTOM consortium. Gateways tothe FANTOM5 promoter level mammalian expression atlas. Genome Biol 2015;16:22

diabetes.diabetesjournals.org McLaughlin and Associates 1697

21. Hald J, Galbo T, Rescan C, et al. Pancreatic islet and progenitor cell surfacemarkers with cell sorting potential. Diabetologia 2012;55:154–16522. Maecker HT, Todd SC, Levy S. The tetraspanin superfamily: molecular fa-cilitators. FASEB J 1997;11:428–44223. Berditchevski F. Complexes of tetraspanins with integrins: more than meetsthe eye. J Cell Sci 2001;114:4143–415124. Zemni R, Bienvenu T, Vinet MC, et al. A new gene involved in X-linkedmental retardation identified by analysis of an X;2 balanced translocation. NatGenet 2000;24:167–17025. Burbelo PD, Ching KH, Mattson TL, Light JS, Bishop LR, Kovacs JA. Rapidantibody quantification and generation of whole proteome antibody response

profiles using LIPS (luciferase immunoprecipitation systems). Biochem BiophysRes Commun 2007;352:889–89526. Ziegler AG, Bonifacio E; BABYDIAB-BABYDIET Study Group. Age-related isletautoantibody incidence in offspring of patients with type 1 diabetes. Diabetologia2012;55:1937–194327. Barker JM, Barriga KJ, Yu L, et al.; Diabetes Autoimmunity Study in theYoung. Prediction of autoantibody positivity and progression to type 1 diabetes:Diabetes Autoimmunity Study in the Young (DAISY). J Clin Endocrinol Metab2004;89:3896–390228. Ziegler AG, Rewers M, Simell O, et al. Seroconversion to multiple islet autoan-tibodies and risk of progression to diabetes in children. JAMA 2013;309:2473–2479

1698 Tetraspanin-7 in Type 1 Diabetes Diabetes Volume 65, June 2016