The Ubiquitin-like Modifier FAT10 Is Selectively Expressed in Medullary Thymic Epithelial Cells and Modifies T Cell Selection

This information is current as Stefanie Buerger, Valerie L. Herrmann, Sarah Mundt, Nico of September 23, 2021. Trautwein, Marcus Groettrup and Michael Basler J Immunol published online 23 September 2015 http://www.jimmunol.org/content/early/2015/09/23/jimmun ol.1500592 Downloaded from

Supplementary http://www.jimmunol.org/content/suppl/2015/09/23/jimmunol.150059 Material 2.DCSupplemental http://www.jimmunol.org/ Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

by guest on September 23, 2021 *average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2015 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Published September 23, 2015, doi:10.4049/jimmunol.1500592 The Journal of Immunology

The Ubiquitin-like Modifier FAT10 Is Selectively Expressed in Medullary Thymic Epithelial Cells and Modifies T Cell Selection

Stefanie Buerger,* Valerie L. Herrmann,* Sarah Mundt,* Nico Trautwein,† Marcus Groettrup,*,‡ and Michael Basler*,‡

HLA-F adjacent transcript 10 (FAT10) is a cytokine-inducible ubiquitin-like modifier that is highly expressed in the thymus and directly targets FAT10-conjugated for degradation by the proteasome. High expression of FAT10 in the mouse thymus could be assigned to strongly autoimmune regulator–expressing, mature medullary thymic epithelial cells, which play a pivotal role in negative selection of T cells. Also in the human thymus, FAT10 is localized in the medulla but not the cortex. TCR Vb-

segment screening revealed a changed T cell repertoire in FAT10-deficient mice. Analysis of five MHC class I– and II–restricted Downloaded from TCR-transgenic mice demonstrated an altered thymic negative selection in FAT10-deficient mice. Furthermore, the repertoire of peptides eluted from MHC class I molecules was influenced by FAT10 expression. Hence, we identified FAT10 as a novel modifier of thymic Ag presentation and epitope-dependent elimination of self-reactive T cells, which may explain why the fat10 could recently be linked to enhanced susceptibility to virus-triggered autoimmune diabetes. The Journal of Immunology, 2015, 195: 000–000. http://www.jimmunol.org/ he maturation of T cells in the thymus is a highly or- Ubiquitin-like modifiers (ULM) posttranslationally modify chestrated process. Somatic TCR gene rearrangements of cellular targets in diverse biological pathways in analogy to the T several distinct gene segments called variable, joining, ubiquitin system. The covalent modification with the cytokine- and diversity leads to unique TCRs expressed on individual inducible ULM HLA-F adjacent transcript 10 (FAT10) tar- T lymphocytes. These developing T cells interact via their TCR gets proteins in a ubiquitin-independent manner for proteasomal with self-peptide–MHC complexes that are displayed by thymic degradation (3–5). FAT10 is conjugated to its substrates via APCs. In a process called positive selection, immature double- isopeptide-linkage mediated by an E1, E2, and possibly E3 positive (DP) thymocytes (CD4+CD8+) that express TCRs with enzyme cascade, in which UBA6 and UBA6-specific E2 en- intermediate avidity for self-peptide–MHC complexes are selected zyme (6, 7) serve as E1-type activating and E2-type conjugating by guest on September 23, 2021 to differentiate into mature single-positive (SP) thymocytes (CD4+ enzymes, respectively (6, 8, 9). The insight into the biological 2 2 CD8 or CD4 CD8+). Selected thymocytes that express TCRs function of FAT10 is still in its infancy. So far, FAT10 has been with high affinity for self-Ags and self-MHC are intrathymically implicated in multiple cellular processes like apoptosis (4, 10), eliminated during negative selection to produce a self-tolerant T cell spindle checkpoint control during mitotic cell cycle (11), and repertoire (1). Mature medullary thymic epithelial cells (mTECs) NF-kB activation (12). Basal FAT10 expression is most prom- contribute to self-tolerance through the promiscuous expression of inent in organs of the immune system, like thymus, fetal liver, tissue-specific Ags in the thymus, which is mainly controlled by the lymph nodes, and spleen (13, 14). In addition, expression of autoimmune regulator (Aire) (2). FAT10 can be synergistically induced by the proinflammatory cytokines IFN-g and TNF-a (15), and it is upregulated during *Division of Immunology, Department of Biology, University of Konstanz, D-78457 dendritic cell (DC) maturation (14). The fusion of FAT10 to the Konstanz, Germany; †Department of Immunology, Interfaculty Institute for Cell N termini of two different viral Ags strongly accelerated their Biology, University of Tubingen,€ D-72076 Tubingen,€ Germany; and ‡Biotechnology Institute Thurgau at the University of Konstanz, CH-8280 Kreuzlingen, Switzerland degradation by the proteasome and enhanced the presentation of their T cell epitopes on MHC class I (MHC-I) molecules Received for publication March 10, 2015. Accepted for publication August 25, 2015. (16–18). These results suggest that linkage to FAT10 can feed This work was supported by German Research Foundation Grant GR 1517/10-2 and the Collaborative Research Center SFB969 (Project C01) (to M.G.). Ags into the MHC-I Ag processing pathway. Moreover, FAT10 Address correspondence and reprint requests to Dr. Michael Basler and Prof. Marcus binds to the autophagy adaptor p62 (19) and colocalizes in Groettrup, Division of Immunology, Department of Biology, University of Konstanz, cells with p62 and the autophagosome marker LC3B (20), Universitaetsstrasse 10, D-78457 Konstanz, Germany. E-mail addresses: michael. which opens the possibility that FAT10-conjugated proteins [email protected] (M.B.) and [email protected] (M.G.) may be processed for MHC class II (MHC-II)–mediated presen- The online version of this article contains supplemental material. tation by targeting them to autophagosomes. Analysis of FAT10- Abbreviations used in this article: Adig, autoimmune regulator–driven Igrp-Gfp; deficient mice demonstrated an enhanced sensitivity toward Aire, autoimmune regulator; cDC, conventional DC; cTEC, cortical thymic 2/2 epithelial cell; DC, dendritic cell; DN, double-negative; DP, double-positive; EpCAM, endotoxin challenge (13). Additionally, FAT10 mice have epithelial cellular adhesion molecule; FAT10, HLA-F adjacent transcript 10; GP, an extended lifespan and reduced adiposity compared with glycoprotein; HEK, human embryonic kidney; HPRT, hypoxanthine-guanine phosphoribosyltransferase; LCMV, lymphocytic choriomeningitis virus; MHC-I, wild-type mice (21). MHC class I; MHC-II, MHC class II; mTEC, medullary thymic epithelial cell; In this study, we could assign the previously recognized high pDC, plasmacytoid DC; SP, single-positive; TEC, thymic epithelial cell; tg, trans- FAT10 expression in the thymus (14) to terminally differenti- genic; ULM, ubiquitin-like modifier. ated mTECs. Thereby, FAT10 influenced the selection of thy- Copyright Ó 2015 by The American Association of Immunologists, Inc. 0022-1767/15/$25.00 mocytes probably due to negative selection and altered peptide

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1500592 2 FAT10 CODETERMINES THYMIC T CELL SELECTION presentation. Hence, we describe a previously unknown function temperature followed by incubation with primary Abs in 5% (w/v) BSA/ of FAT10 in modifying thymic T cell selection. PBS overnight at 4˚C. After washing three times with PBS/0.1% (v/v) Tween 20, the secondary Abs were diluted in 5% (w/v) BSA/PBS and incubated for 1 h at room temperature. Sections were mounted using Materials and Methods Kaiser’s glycerol gelatin (Merck) or Mowiol. All images were obtained Mice, viruses, and media using a point laser-scanning microscope Zeiss LSM 510 Meta (Carl Zeiss) at the Bioimaging Center of the University of Konstanz. Image analysis C57BL/6 mice (H-2b) were originally purchased from Charles River 2/2 was performed using ImageJ software (National Institutes of Health, Laboratories (Sulzfeld, Germany). FAT10 mice (13) were kindly pro- Bethesda, MD). The following reagents were used: monoclonal mouse vided by A. Canaan and S. M. Weissman (Yale University School of anti-human FAT10 (4F1) was purified from hybridoma supernatant Medicine, New Haven, CT). P14 mice (transgenic [tg] line 318) (22) were (19). Rat anti-human Aire (TM-724) was purchased from eBioscience obtained from Dr. Oliver Planz (Tubingen€ University, Tubingen,€ Ger- (Frankfurt, Germany). Polyclonal rabbit anti-Cytokeratin 5 was purchased many). OT-I (23), OT-II (24), and Smarta mice (25), were provided by the from Covance (purchased via Hiss Diagnostic, Freiburg, Germany). Mu- Swiss Immunological Mutant Mouse Repository. HY TCR-tg mice (26) rine thymic cortex was stained with mouse anti-cytokeratin 18 (RGE53; were obtained from A. Tafuri (Deutsches Krebsforschungszentrum, Hei- Thermo Scientific) or mouse anti-cytokeratin 8 Ab (Ks8.7; Progen, Hei- delberg, Germany). Aire-driven Igrp-Gfp (Adig) mice (27) were kindly delberg, Germany). Secondary Abs were donkey anti-rabbit Cy3, goat anti- provided by Ksenija Jovanovic (Ludwig-Maximilians-Universita¨t mouse Alexa Fluor 488, and goat anti-rat Alexa Fluor 633 (all from Life Munchen,€ Munich, Germany) with the kind permission of Mark S. Technologies). Anderson (University of California, San Francisco, San Francisco, CA). Mice were kept in a specific pathogen-free facility. Sex- and age-matched Flow cytometry and cell preparation mice were used at 6–10 wk of age. Animal experiments were approved by the review board of Regierungspra¨sidium Freiburg. HEK293 cells Flow cytometry was performed exactly as previously described (29). To isolate different thymus-derived cell populations, the following procedure (American Type Culture Collection) and stably FLAG-FAT10–expressing Downloaded from HEKFAT10 cells (20) were grown in IMDM containing GlutaMAX was used: thymi from 1–3-wk-old mice were cut into small pieces and (Invitrogen-Life Technologies, Karlsruhe, Germany) supplemented with digested at 37˚C in RPMI 1640 containing 0.25 mg/ml collagenase D 10% FCS, 100 U/ml penicillin, and 100 mg/ml streptomycin. (Roche), 1 U/ml dispase I (Life Technologies), 2% (v/v) FBS (Life Lymphocytic choriomeningitis virus (LCMV)-clone 13 was obtained Technologies), 25 mM HEPES (pH 7.2), and 25 mg/ml DNase I (Roche), from M. van den Broek (Zurich, Switzerland) and propagated on the followed by incubation for 5 min in 5 mM EDTA. Cells were washed and 7 r fibroblast line L929. LCMV carrier mice were obtained by injecting 10 resuspended in Percoll ( , 1.115; Sigma-Aldrich). A discontinuous gradi- PFU LCMV clone 13 i.p. into newborn mice (within 24 h after birth). ent was then generated by the addition of a layer of Percoll (r, 1.065) followed by a layer of 13 PBS on top of this cell suspension. Gradients http://www.jimmunol.org/ Quantitative real-time RT-PCR were spun for 30 min at 1.350 3 g at 4˚C. Thymocytes were collected from the lower interface (high-density fraction) and stained for sorting by RNA was prepared using the RNeasy Plus Micro Kit (Qiagen, Hilden, flow cytometry. Low-density cells containing DCs, macrophages, TECs, Germany) according to the manufacturer’s instructions. For the synthesis of and fibroblasts were collected from the upper interface. CD11c+ DCs were single-stranded cDNA from pure total RNA, the Sensiscript RT kit (Qia- isolated by MACS using mouse CD11c microbeads (Miltenyi Biotec, gen) was used. To determine relative , the LightCycler Bergisch Gladbach, Germany) and further subdivided into plasmacy- Fast Start DNA Master SYBR Green I Kit was used in conjunction with toid DCs (pDCs; CD11cmidCD45R/B220+), migratory conventional DCs the Light-Cycler Instrument and the LightCycler Software Version 3.5 (cDCs; CD11chiSIRPa+), or autochthonous (CD11chiSIRPa2) thymic DCs (all from Roche, Mannheim, Germany) according to the manufacturer’s by FACS. Remaining stromal TECs were washed and stained for FACS instructions. The following primers were used: FAT10 forward, 59- (mTECs: CD452epithelial cellular adhesion molecule [EpCAM]+Ly-512; GGGATTGACAAGGAAACCACTA-39 and FAT10 reverse, 59-TTCA- cTECs: CD452EpCAM+Ly-51+). by guest on September 23, 2021 CAACCTGCTTCTTAGGG-39; Aire forward, 59-GTCCCTGAGGACAA- Thymocytes were screened for TCR Vb expression by flow cytometry GTTCCA-39 and Aire reverse, 59-TCTTTGAGGCCAGAGTTGCT-39; using PE-conjugated CD4 (GK1.5) and allophycocyanin-conjugated Psmb11 forward, 59-GACCATCCAGGAAGCCTACA-39 and Psmb11 CD8 (53-6.7) Abs (both from eBioscience) in conjunction with a Vb reverse, 59-ATACAGCACGCAAGCATCAC-39; GAPDH forward:, 59- TCR screening panel (BD Biosciences, Heidelberg, Germany) according GTGTTCCTACCCCCAATGT-39 and GAPDH reverse, 59-TGTCATCA- to the manufacturer’s instructions. Intracellular Foxp3 staining was TACTTGGCAGGTTTC-39; and hypoxanthine-guanine phosphoribosyl- performed with the anti-mouse Foxp3 mAb FJK-16s according to stan- transferase (HPRT) forward, 59-CCAGCAGGTCAGCAAAGAACTTA-39 dard procedures. and HPRT reverse, 59-TGGACAGGACTGAAAGACTTG-39. The runs The following Abs were used for flow cytometry: FITC-conjugated were programmed as follows: denaturation for 10 min at 95˚C, amplifi- anti-CD4 (H129.19), PE-conjugated anti-CD4 (RM4-5), PE- or allophycocyanin- cation (40 cycles with reading of the fluorescence at the end of each cycle) conjugated anti-CD11c (HL3), PerCP-conjugated anti-CD45R/B220 for 10 s at 95˚C, 10 s at annealing temperature, 10 s at 72˚C; and analysis (RA-3-6B2), PE-Cy7–conjugated anti-CD45 (30-F11), PE-conjugated of the products (reading of the fluorescence in a continuous mode) for 0 s anti-Ly-51 (BP-1), FITC-conjugated anti-CD80 (16-10A1), FITC-conjugated at 95˚C, 62 to 95˚C transition with 0.1˚C increment/s. The specificity of the anti-CD172a/SIRPa (P84), FITC-conjugated anti–H2-IAb (AF6-120.1), amplification was verified by melting curve analysis (95˚C for 0 s; 65˚C for V500-conjugated anti–I-A/I-E (M5/114.15.2), FITC-conjugated anti–H2-Kb 15 s, and 95˚C for 0 s with a temperature transition rate of 0.1˚C/s in (AF6-88.5), FITC-conjugated anti–H2-Db (KH95), and PE-conjugated a continuous acquisition mode). Relative gene expression was normalized anti-Va2 TCR (B20.1) Abs were purchased from BD Biosciences. to HPRT or GAPDH content and evaluated according to the Pfaffl method Allophycocyanin-conjugated anti-CD326/EpCAM (G8.8) was purchased using the Excel-based software tool REST-384-b (28). from BioLegend (Fell, Germany). Dead cells were excluded by DAPI (Sigma-Aldrich) staining. Histological analysis Peptide elution and mass spectrometry Human thymic tissue samples originated from pediatric patients under- going routine cardiac surgery and were a kind donation of Prof. Bruno HLA ligands of HEK293 and HEKFAT10 cells were isolated by immu- Kyewski (German Cancer Research Center, Heidelberg, Germany). Murine noaffinity chromatography. Cells were lysed in buffer containing PBS, thymus was retrieved from 3-wk-old animals. Thymic tissue was embedded 0.6% CHAPS, and complete protease inhibitor (Roche), shaken for 1 h, and in Tissue-Tec OCT (Sakura Finetek, Staufen, Germany), snap-cap frozen in subsequently sonicated for 1 min. Following centrifugation for 1.5 h to liquid nitrogen, and stored at 280˚C until use. Sections were cut at 5-mm remove debris, the supernatant was applied on affinity columns overnight. thickness using an 2800 Frigocut instrument (Reicher-Jung, Heidelberg, Columns were previously prepared by coupling Abs to CNBr-activated Germany) at 225˚C in conjunction with a knife in C-cut cooled to 220˚C. Sepharose (GE Healthcare, Buckinghamshire, U.K.) (1 mg Ab/40 mg Sections were mounted on Super Frost Plus slides (Thermo Scientific, Sepharose). Ab W6/32 was used to isolate HLA class I molecules. On the Ulm, Germany), fixed in cold acetone (Sigma-Aldrich), air dried, and following day, the columns were eluted in eight steps using 0.2% tri- stored at 220˚C until further use. For staining procedure, sections were fluoroacetic acid. Filtration of the eluate through a 10-kDa filter (Merck dried at room temperature for 1 h and circled with a liquid blocker. All Millipore, Darmstadt, Germany) yields the HLA ligands in solution. The incubation steps of the subsequent staining procedures were performed in filtrate was desalted with C18 ZipTips (Merck Millipore) and subsequently a humidified chamber. Sections were permeabilized with PBS/0.1% (v/v) concentrated using a vacuum centrifuge (Bachofer, Munchen,€ Germany). Tween 20 (Sigma-Aldrich) for 10 min at room temperature. Blocking was Sample volume was adjusted for measurement by adding 1% acetonitrile/ performed with 5% (w/v) BSA (Sigma-Aldrich)/PBS for 30 min at room 0.05% trifluoroacetic acid (v/v). With an injection volume of 5 ml, HLA The Journal of Immunology 3 Downloaded from http://www.jimmunol.org/ by guest on September 23, 2021 4 FAT10 CODETERMINES THYMIC T CELL SELECTION

FIGURE 2. FAT10 deficiency in mTECs alters neither MHC surface expression nor mTEC maturation. (A) Flow cytometry histograms of H2-Kb, H2-Db, and H2-IAb sur- face expression on mTECs (CD452EpCAM+Ly-512) (top panel) and cTECs (CD452EpCAM+Ly-51+)(bottom panel) derived from C57BL/6 (red) or FAT102/2 (blue) mice. As a negative control, cells were not stained with the respective MHC Ab (fluorescence-minus-one [FMO]). One representative experiment out of three is shown. (B) mTECs (CD452EpCAM+Ly-51) derived from C57BL/6 (red) and FAT102/2 (blue) mice were analyzed by flow cytometry for the expression of the surface markers CD80 and MHC-II. mTEClo and mTEChi cells were discrimi- natedbyCD80expression(left panel) and analyzed for b H2-IA surface expression (middle and right panels). One Downloaded from representative experiment out of three is shown. http://www.jimmunol.org/ ligands were loaded (100 mm 3 2 cm, C18, 5 mm, 100 A˚ ) and separated (version 6.04) (GraphPad, San Diego, CA). Statistical significance was (75 mm 3 50 cm, C18, 3 mm, 100 A˚ ) on Acclaim Pepmap100 columns achieved when p , 0.05 (*p , 0.05, **p , 0.01, ***p , 0.001). (Dionex, Sunnyvale, CA) using an Ultimate 3000 RLSCnano uHPLC system (Dionex). A gradient ranging from 2.4–32% acetonitrile/H2O with 0.1% formic acid was used to elute the peptides from the columns over 140 min at Results a flow rate of 300 nl/min. Online electrospray ionization was followed by FAT10 is highly expressed in mTECs tandem mass spectrometry analysis in a LTQ Orbitrap XL instrument (Thermo Fisher Scientific, Bremen, Germany). Survey scans were acquired FAT10 expression is mainly restricted to lymphoid organs. To in the Orbitrap mass analyzer with a resolution of 60,000 and a mass range determine which cell type is responsible for the particularly strong by guest on September 23, 2021 of 400–650 mass-to-charge ratio. Peptides with a charge state other than 2+ FAT10 expression in the thymus, autochthonous cDCs (CD11chi or 3+ were rejected from fragmentation. Fragment mass spectra of the five SIRPa2), migratory cDCs (CD11chiSIRPa+), pDCs (CD11cmid most intense ions of each scan cycle were recorded in the linear ion trap 2 CD45R+), CD4 SP thymocytes (CD4+CD8 ), CD8SP thymocytes (top5 CID). Normalized collision energy of 35%, activation time of 30 ms, 2 + 2 2 and isolation width of 2 mass-to-charge ratio was used for fragment mass (CD4 CD8 ), double-negative (DN) thymocytes (CD4 CD8 ), analysis. Dynamic exclusion was set to 1 s. The RAW files were processed and DP thymocytes (CD4+CD8+) were purified by FACS and against the human proteome as comprised in the Swiss-Prot database (http:// analyzed for FAT10 mRNA expression by real-time RT-PCR www..org; status: December 12, 2012; 20,225 reviewed sequences (Fig. 1A). Compared to the total thymic FAT10 mRNA expres- contained) using MASCOT server version 2.3.04 (Matrix Science, Boston, MA) and Proteome Discoverer 1.4 (Thermo Fisher Scientific). Oxidation of sion, only low FAT10 expression could be observed in these cell methionine was allowed as dynamic peptide modification. A mass tolerance populations. Therefore, FAT10 mRNA expression was determined of 5 ppm or 0.5 Da was allowed for parent and fragment masses, respec- in mTECs and cTECs by real-time RT-PCR (Fig. 1B). Remark- tively. Filtering parameters were set to a Mascot Score ,20, search engine ably, a very high FAT10 mRNA expression could be found in rank = 1, peptide length of 8–12 aa, achieving a false discovery rate of 5% as determined by an inverse decoy database search. mTECs but not in cTECs. The purified mTEC population was at least 80% pure according to flow cytometry with mRNA expres- Statistical analysis sion for Aire being largely restricted to mTECs and expression of The statistical significance was determined using the Student t test. All the mRNA for the thymoproteasome subunit b5t (Psmb11) (30) statistical analyses were performed using GraphPad Prism software being confined to cTECs. The residual FAT10 mRNA expression

FIGURE 1. High FAT10 expression in terminally differentiated mTECs. (A)CD42CD82 (DN), CD4+CD8+ (DP), CD4+CD82 (CD4SP), and CD42CD8+ (CD8SP) thymocytes as well as thymic pDCs (CD11cmidCD45R+), thymic migratory cDCs (CD11chiSIRPa+), and thymic autochthonous cDCs (CD11chi SIRPa2) were purified by FACS, and FAT10 mRNA expression was determined by real-time RT-PCR and compared with total thymus tissue. Thymocyte and regulatory T cell distribution in C57BL/6 and FAT102/2 mice are shown in the related Supplemental Fig. 3. (B) DN, DP, CD4SP, and CD8SP thymocytes as well as thymic DCs (CD11c+), mTECs (CD452EpCAM+Ly-512), and cTECs (CD452EpCAM+Ly-51+) were purified by FACS, and FAT10 mRNA was quantified. Representative flow cytometry profiles of the sorting strategy for mTECs and cTECs are depicted in the top panel.(C) TECs from Adig (Aire reporter) mice were isolated by FACS according to the surface markers CD452EpCAM+Ly-512 (mTECs) and CD452EpCAM+Ly-51+ (cTECs) with mTECs being further divided into mTEClo (MHC-IIlo), mTEChi (MHC-IIhi), and mTEChi (MHC-IIhi)Aire+. Thymocytes were used as negative control. FAT10, Aire (B and C), and PSMB11 (B and C) mRNA levels were determined by real-time RT-PCR and calculated relative to the total thymus and normalized to HPRT (A and B)orGAPDH(C). Columns represent mean values 6 SEM of two (A and C) or three (B) independent experiments. (D) Total thymic FAT10 and Aire mRNA levels of FAT102/2, Aire2/2, and wild-type mice were determined by real-time RT-PCR, calculated relative to the wild-type thymus, and normalized to GAPDH. (E) Five-micrometer cryosections of human pediatric patients were costained with mouse anti-human FAT10 (clone 4F1) and rabbit anti-Keratin 5 followed by secondary Abs goat anti-mouse Alexa 488 and donkey anti-rabbit Cy3 and analyzed by confocal microscopy. Patient numbers are indicated on the right. One representative image out of three independent experiments is displayed. Scale bar, 100 mm. SSC, side scatter. The Journal of Immunology 5 in cTECs is comparable to that of Aire, suggesting that it stems the TCR (Fig. 3A, 3B). When comparing SP thymocytes from from few contaminating cTEC cells. It therefore appears that wild-type and FAT10-deficient mice significant alterations in Vb- FAT10 is expressed in mTECs and not or only negligibly in segment usage were recorded for TCR-Vb10b and TCR-Vb12 in cTECs. Furthermore, immunohistological analysis of thymus cryo- CD8SP thymocytes and for Vb3, Vb5.1/2, Vb6, Vb10b, Vb11, sections from wild-type or FAT10-deficient mice revealed no al- Vb12, and Vb14 for CD4SP T cells. Nevertheless, analysis of Vb- teration in medulla and cortex structure (Supplemental Fig. 1). segment usage of DP cells revealed no difference between wild- According to the terminal differentiation model (31), mTECs type and FAT10-deficient mice (Supplemental Fig. 3D), strongly undergo a maturation program that goes along with upregulation suggesting that FAT10 influences negative selection. Hence, it of MHC-II and costimulatory molecules like CD80 and the ex- appears that FAT10 expression modifies T cell repertoire selection pression of Aire. Immature mTECs express low (lo) levels of both in the thymus. lo lo lo surface molecules and are thus termed mTEC (CD80 MHC-II ). 2 2 Thymic T cell selection is altered in FAT10 / mice Upon maturation, mTECs differentiate to mTEChiAire2 cells (CD80hiMHC-IIhi) and terminally differentiate into mTEChiAire+ According to TCR-Vb usage analysis minor differences in bulk cells. To determine FAT10 mRNA expression in these three sub- T cell selection in the thymus of FAT10-deficient mice were ob- sets of mTECs, the different mTEC populations were isolated served (Fig. 3). To investigate how apparent the effect of FAT10 from Aire reporter (Adig) mice by FACS according to the markers deficiency might be on the level of single TCRs, different TCR-tg 2/2 mentioned above (Fig. 1C). Real-time RT-PCR analyses of these C57BL/6 mice were crossed to FAT10 mice on the same 2/2 cell populations demonstrated strongly elevated FAT10 expres- background. CD4SP and CD8SP thymocytes of P14 FAT10 b sion in terminally differentiated Aire expressing mTECshiAire+.The (TCR tg for the LCMV-WE glycoprotein (GP)33–41/H-2D ), Downloaded from 2/2 b sorting efficiency exceeded 80% for all three populations. In order OT-I FAT10 (TCR-tg for OVA257-264/H-2K ), and female 2/2 to determine whether FAT10 expression in the thymus is regulated HY FAT10 (TCR-tg for the male HY epitope KCSRNRQYL/ b by Aire, FAT10 mRNA expression was analyzed in Aire-deficient H-2D )wereanalyzedbyflowcytometry(Fig.4A).Compared 2/+ 2/2 mice. FAT10 expression in the thymus was not altered in Aire- to P14 FAT10 mice, P14 FAT10 mice showed an ∼40% deficient mice (Fig. 1D), indicating Aire-independent FAT10 ex- reduction (from 3.8 to 2.3%) in CD8SP thymocytes, whereas

pression in the thymus. CD4SP were not affected. The percentage of CD8SP using http://www.jimmunol.org/ To determine if also FAT10 expression is restricted to the the tg TCR (Vb8.1/2) was slightly but significantly reduced medulla and to test if medullary FAT10 expression is also valid in in P14 mice lacking FAT10. Similarly, comparison of OT-I 2/+ 2/2 humans, we used a highly specific anti-human FAT10 mAb (19) FAT10 with OT-I FAT10 mice revealed an ∼25% re- for immunohistological analysis of cryosections from three human duction (from 10.4 to 7.8%) in CD8SP thymocytes with no thymi (Fig. 1E). Consistent with the FAT10 mRNA expression in alteration in CD4SP thymocytes. The usage of the tg TCR the mouse, FAT10 strongly colocalized with the medulla-specific (Va2) of CD8SP was comparable between FAT10-deficient marker keratin 5 in all three human thymus samples, indicating and FAT10-proficient OT-I mice. FAT10-deficient female HY FAT10 protein expression in the medulla but not the cortex. Fur- thermore, both FAT10 and Aire expression is restricted to the by guest on September 23, 2021 medulla (Supplemental Fig. 2). Lack of FAT10 alters neither MHC expression on mTECs nor mTEC maturation In order to investigate whether the high FAT10 expression in mTECs (Fig. 1B, 1C) influences MHC surface expression, the MHC-I (H2-Kb and H2-Db) and MHC-II (H2-IAb) expression on mTECs and cTECs of C57BL/6 wild-type and FAT102/2 mice was analyzed by flow cytometry (Fig. 2A). No difference in MHC- I and MHC-II surface expression could be detected between wild- type and FAT10-deficient mTECs and cTECs. Additionally, the lack of FAT10 did not alter H2-IAb expression on mTEClo and mTEChi (Fig. 2B). Because MHC-II molecules are upregulated during mTEC maturation and serve as maturation markers, these results strongly suggest that the maturation of mTECs proceeds normally in the absence of FAT10. FAT10 deficiency causes changes in the T cell repertoire To investigate whether FAT10 expression in mTECs affects T cell selection, thymocyte subsets derived from wild-type or FAT0- deficient mice were analyzed at different stages of thymic selec- tion by flow cytometry (Supplemental Fig. 3A, 3B). No difference either in relative or absolute numbers of DN, DP, CD4SP, and CD8SP thymocyte populations could be observed. Additionally, stainings for CD4 and intracellular Foxp3 revealed that FAT10 + deficiency did not alter the number of Foxp3 regulatory T cells in FIGURE 3. An altered T cell repertoire in FAT10-deficient mice. Flow the thymus (Supplemental Fig. 3C). cytometric analysis of indicated Vb variable segments of TCRs from To investigate whether FAT10 expression alters the T cell rep- CD8SP (A) or CD4SP (B) thymocytes derived from C57BL/6 or FAT102/2 ertoire in the thymus, CD4SP and CD8SP cells in the thymus were mice. Graphs show the mean 6 SEM of two independent experiments screened with a panel of Abs specific for different Vb-segments of (n =10).*p , 0.05, **p , 0.01, ***p , 0.001. 6 FAT10 CODETERMINES THYMIC T CELL SELECTION Downloaded from http://www.jimmunol.org/ by guest on September 23, 2021

FIGURE 4. Altered thymic selection in different TCR-tg FAT10-deficient mice. (A) Thymocytes derived from FAT102/+ and FAT102/2 HY female (f), OT-I, and P14 TCR-tg mice were stained for CD4, CD8, and TCR-Vb8.1/2 (HY, P14) or Va2 (OT-I), respectively, and analyzed by flow cytometry. Graphs show the mean percentages 6 SEM (HY FAT102/+: n =25;HYFAT102/2: n =22;P14FAT102/+: n =57;P14FAT102/2: n = 53; OT-I FAT102/+: n = 40; OT-I FAT102/2: n = 32). Representative flow cytometry dot plots are shown in the left panel.(B) Peripheral blood was stained for CD8 and analyzed by flow cytometry. Shown are the mean percentages 6 SEM of CD8 of lymphocytes (HY FAT102/+: n =36;HY FAT102/2: n = 50; P14 FAT102/+: n = 59; P14 FAT102/2: n = 62; OT-I FAT102/+: n = 37; OT-I FAT102/2: n = 31). (C) Thymocytes derived from FAT102/+ and FAT102/2 OT-II and Smarta TCR-tg mice, respectively, were stained for CD4, CD8, and Va2 and analyzed by flow cytometry. Graphs show the mean percentages 6 SEM (OT-II FAT10+/+: n = 40; OT-II FAT102/2: n = 34; Smarta FAT10+/+: n = 23; Smarta FAT102/2: n = 34). Representative flow cytometry dot plots are shown in the left panel.(D) Peripheral blood was stained for CD4 and analyzed by flow cytometry. Shown (Figure legend continues) The Journal of Immunology 7

TCR-tg mice showed no differences in CD4SP and CD8SP female HY mice, selection proceeds normally, whereas in male thymocytes compared with wild-type female HY TCR-tg mice. mice, presentation of the HY peptide on mTECs leads to elimi- No difference in absolute thymocyte number between wild-type nation of the majority of clonotypic thymocytes at the DN to DP and FAT102/2 could be observed in the analyzed TCR-tg mice transition. Therefore, in male TCR-tg mice, a strong reduction in (Supplemental Fig. 3E). Because not all TCRs were affected by DP thymocytes and CD8 tg cells could be observed, whereas in FAT10 deficiency, it appears that FAT10 influences the selec- female mice, tg cells are positively selected to the CD8 lineage tion of thymocytes in a TCR-selective manner. Similar to the (Fig. 4A) (26, 32). The percentage of DP (data not shown) and thymus, CD8+ cells in the peripheral blood of P14 and OT-I CD8SP thymocytes (Figs. 4A, 5C), both in male HY and male HY FAT10-deficient mice were significantly reduced, whereas FAT102/2, was, compared with female mice, markedly reduced, CD8+ cells in FAT10-deficient female HY mice were not af- demonstrating the deletion of TCR-tg cells in this model. No fected (Fig. 4B). difference of DP was observed between male HY FAT10-deficient In order to investigate whether FAT10 alters thymic selection in and -proficient mice. Compared to male HY FAT102/+ mice, MHC-II–restricted TCR-tg mice, OT-II (TCR-tg for OVA323-339/ CD8SP cells in the thymus were strongly reduced (by ∼40%) in H-2IAb) and Smarta mice (TCR-tg for the LCMV-WE GP61–80/ male HY FAT102/2 mice (from 2.2 to 1.3%), whereas CD4SP H-2IAb) were crossed to FAT102/2 mice on the C57BL/6 back- cells were not affected. Because negative selection occurs at the ground. OT-II FAT102/2 mice showed an increased percentage of DP to CD8SP transition, one can be confident that the reduced CD4SP thymocytes and a reduced percentage of CD8SP thymo- percentage of CD8SP in male HY FAT102/2 mice is based on cytes compared with OT-II FAT10-proficient mice (Fig. 4C). No enhanced negative selection in FAT102/2 mice. difference for CD4SP could be observed between Smarta TCR-tg Downloaded from Reduced survival of neonatally LCMV-infected FAT10-deficient and FAT10-proficient mice, whereas CD8SP were FAT10-deficient TCR-tg Smarta mice increased in Smarta FAT102/2 compared with Smarta FAT10+/+ + mice (Fig. 4C). CD4+ T cells in the peripheral blood were not In order to study negative selection of CD4 cells, the LCMV-carrier altered in OT-II TCR-tg mice, whereas the percentage of CD4+ model of negative selection was further studied in Smarta TCR-tg T cells was significantly reduced in the blood of Smarta TCR-tg mice. Neonatal infection with LCMV clone 13 led to a strong 2/2 FAT10 compared with Smarta wild-type mice (Fig. 4D). mean reduction of CD4SP cells from 48.2 to 28.0% of thymocytes http://www.jimmunol.org/ and a corresponding enhancement of CD8SP cells, indicating that Enhanced negative selection in the thymi of FAT10-deficient negative selection of tg LCMV-GP61–80/H-2IAb–specific T cells MHC-I–restricted male HY and LCMV carrying P14 TCR-tg occurred (Fig. 5E). During these experiments, only a low percent- mice age of neonatally LCMV-infected Smarta mice survived (Table I). High FAT10 expression in the thymus was restricted to mTECs Whereas 7 tg mice out of 33 FAT10-proficient Smarta TCR-tg (Fig. 1). Because mTECs are mainly responsible for negative carrier litters survived, only 1 out of 31 Smarta FAT102/2 carrier selection of thymocytes, we decided to study negative selection in mice remained viable. Remarkably, although the percentage mouse models for negative selection. Mice infected at birth with of TCR-Va2+ of CD4SP T cells in the seven analyzed FAT10-

LCMV (referred to as carrier mice) grow to adulthood with high proficient Smarta carrier mice decreased from 97.3% in noncar- by guest on September 23, 2021 viral loads persisting in nearly all tissues. Viral Ags presented in rier to 15.4% in carrier mice, the only surviving FAT10-deficient the neonatal thymus cause deletion of virus-specific T cells. Smarta LCMV carrier mouse we could analyze displayed 50.3% tg Hence, the neonatal infection of P14 TCR-tg mice leads to TCR-Va2+ CD4SP thymocytes (Fig. 5E). This result is consistent a clonal deletion of P14 LCMV-specific T cells. Indeed, P14 mice with a lack of negative selection in this class II–restricted model, infected at birth have markedly reduced numbers of CD4SP and which might have led to the low survival rate of FAT10-deficient CD8SP thymocytes and CD8+ peripheral T cells (22). In order Smarta LCMV carrier mice. Due to the low numbers of survivors, to investigate negative selection in this model, P14 and P14 a further analysis could not be performed despite extensive breed- FAT102/2 mice were infected at birth with the fast replicating ing efforts. Nevertheless, the reduced survival of FAT10-deficient LCMV clone 13 strain. At 2 mo of age, the carrier state of all mice Smarta TCR-tg carrier mice compared with wild-type Smarta TCR- was confirmed by determining LCMV titers in the spleen (data not tg mice suggests that FAT10 plays a crucial role in determining shown), and thymocytes were analyzed at different stages of de- survival of LCMV-carrier mice in this tg model. velopment (Fig. 5A). Compared to noncarrier mice (Fig. 4A), the usage of the tg TCR (Vb8.1/2) was reduced from ∼70 to 30% FAT10 expression affects the repertoire of MHC-I–bound (Fig. 5B), demonstrating negative selection of the tg TCR in peptides carrier mice. The percentage of CD8SP in the thymus of P14 Having found exclusive FAT10 expression in mTECs (Fig. 1B) and FAT102/2 compared with P14 wild-type mice was reduced by an altered T cell selection in FAT10-deficient mice (Figs. 3–5), we ∼30% (from 2.9 to 2.1%) (Fig. 5A). CD4SP were not altered in decided to investigate the repertoire of peptide ligands bound to P14 FAT102/2 carrier mice (Fig. 5A). Nevertheless, the analysis MHC-I molecules in the presence and absence of FAT10. Because of the tg TCR (Vb8.1/2) of CD8SP cells in the thymus and CD8+ mTECs cannot be isolated in sufficient number, we used human cells in the spleen revealed an increased negative selection in embryonic kidney (HEK) 293 cells for this analysis, which do not P14 FAT102/2 carrier mice compared with P14 FAT10-proficient express endogenous FAT10 in the absence of stimulation with carrier mice (Fig. 5B). Hence, these results indicate that FAT10 IFN-g/TNF-a (7). An HEK293 cell line stably overexpressing expression alters negative selection in this model. FAT10 (HEKFAT10) (20) was used as a source for class I ligands Next, we investigated thymic T cell selection in HY TCR-tg and compared with wild-type HEK293 cells. To identify naturally mice. These mice express MHC-I–restricted TCR a-and processed peptides, we isolated HLA molecules from HEK293 or b-chains that recognize the male HY Ag presented by H-2Db.In HEKFAT10 cells, eluted the peptides, fractionated them by HPLC,

are the mean percentages 6 SEM of CD4+ cells of lymphocytes (OT-II FAT10+/+: n = 80; OT-II FAT102/2: n = 111; Smarta FAT10+/+: n = 99; Smarta FAT102/2: n = 122). *p , 0.05, **p , 0.01, ***p , 0.001. 8 FAT10 CODETERMINES THYMIC T CELL SELECTION Downloaded from http://www.jimmunol.org/ by guest on September 23, 2021

FIGURE 5. Altered negative selection in P14 and Smarta TCR-tg LCMV carrier and male HY TCR-tg FAT10-deficient as compared with FAT10- proficient mice. (A) Thymocytes derived from wild-type and FAT102/2 P14 LCMV carrier mice were stained for CD4 and CD8 and analyzed by flow cytometry. Graphs show the mean percentages 6 SEM of CD4SP or CD8SP of thymocytes (P14: n = 38; P14 FAT102/2: n = 48). (B) Thymocytes (left panel) and splenocytes (right panel) were stained for CD8, CD4, and Vb8.1/2 and analyzed by flow cytometry. Shown are the mean percentages 6 SEM of Vb8.1/2 of CD8SP (thymus) or CD8+ cells (spleen) (P14: n = 38; P14 FAT102/2: n = 48). (C) Thymocytes derived from FAT102/+ and FAT102/2 male (m) HY mice were stained for CD4, CD8, and Vb8.1/2 and analyzed by flow cytometry. Graphs show the mean percentages 6 SEM (HY FAT102/+ m: n = 26; HY FAT102/2 m: n = 20). (D) Peripheral blood was stained for CD8 and analyzed by flow cytometry. Shown are the mean percentages 6 SEM of CD8+ cells of lymphocytes (HY FAT102/+ m: n = 31; HY FAT102/2 m: n = 29). (E) Thymocytes from uninfected and neonatally LCMV-infected carrier FAT10+/+ 2 2 and FAT10 / Smarta TCR-tg mice as indicated were stained and analyzed for CD4SP (left panel), CD8SP (middle panel), and (Figure legend continues) The Journal of Immunology 9

Table I. Survival rate of Smarta and Smarta FAT102/2 LCMV carrier mice

Total No. of Transgenic Survivor per Litter Litters Survivors (Mean 6 SEM) Smarta 33 7 0.21 6 0.10 Smarta FAT102/2 31 1 0.03 6 0.03 and analyzed the peptides by mass spectrometry. Valid peptide sequences identified with high confidence were compared be- tween samples derived from HEK293 and HEKFAT10 cells FIGURE 6. Identification of peptide ligands eluted from HLA molecules of wild-type (WT) and stably FAT10-transfected HEK293 cells reveals (Supplemental Table I). In three different experiments, we identified a FAT10-mediated alteration of the peptide repertoire. MHC-I–peptide 241 different peptides present in at least two out of three samples complexes were isolated by immunoprecipitation of HLA molecules, and derived from both cell types (Fig. 6). Seven peptides were exclu- MHC-I–eluted peptides were identified by mass spectrometry. Venn dia- sively found in cells overexpressing FAT10, whereas 42 peptides grams show numbers of identified peptides that were unique to WT were only present in HEK293 wild-type cells. Although most of the HEK293 cells, unique to HEKFAT10 cells, or shared by both. The iden- validated peptides were shared by the two strains, we identified tified peptides are listed in Supplemental Table I. a proportion of peptides that were found only in one strain. The Downloaded from results showed that FAT10 expression alters the repertoire of pep- proteasome (3, 5, 34). The proteasome plays a critical role in the tides bound to MHC-I, which most likely accounts for the observed generation of ligands from intracellular Ags for class I presenta- differences in thymic selection in the absence of FAT10. tion. Therefore, it seems likely that FAT10-mediated proteasomal degradation might influence MHC-I Ag presentation. Indeed, an Discussion effect of FAT10 on Ag presentation has been demonstrated in two Cell-type selective expression can be an eye opener for function. independent studies (16, 17). Firstly, the N-terminal fusion of the http://www.jimmunol.org/ In this study, we could assign the high FAT10 expression in the human CMV-derived pp65 Ag to FAT10 facilitated direct MHC-I– thymus to terminally differentiated mTECs. mTECs are critical in restricted presentation and DC-mediated cross-presentation of the establishing and maintaining the appropriate microenvironment for HLA-A2–restricted pp65 (495–503) epitope (17). Secondly, the negative selection and maturation of T cells (33). It is striking long-lived nucleoprotein of LCMV could be targeted for degra- that FAT10 is upregulated during the maturation both of mTECs dation by N-terminal fusion to FAT10 and, in a vaccination setup, (Fig. 1C) and DCs (14), as both are involved in negative selection, induced strong T cell responses (16). Hence, due to its role in although they derive from completely different lineages. Consis- proteasomal degradation, FAT10 in mTECs might influence Ag tent with the notion of a role of FAT10 in negative selection presentation. We compared MHC-I ligands presented on the

mediated by mTECs, FAT10-deficient mice displayed an altered cell surface between HEK293 cells either expressing or lacking by guest on September 23, 2021 T cell repertoire (Fig. 3). Additionally, in FAT10-deficient P14 FAT10. In three independent analyses, we identified 241 class I and OT-I TCR-tg mice, the selection process to CD8SP was not as ligands shared by both cell types (Fig. 6, Supplemental Table I). A efficient as in FAT10-proficient mice, whereas HY TCR-tgs were total of 42 peptides were only found in cells lacking FAT10, not affected (Fig. 4A). This indicates that FAT10 influences T cell whereas 7 peptide sequences were exclusively identified in selection in an epitope-specific manner. In accordance with the FAT10-expressing cells. Hence, it appears that FAT10 influences altered T cell repertoire and the reduced number of CD8SP in two the MHC-I peptidome. TCR-tg mice, negative selection was altered in two mouse models How could FAT10 influence the proteasomal targeting in mTECs for T cell selection (Fig. 5). In P14 FAT102/2 carrier mice, an in- and alter negative selection? A recent proteomic analysis of ULM creased negative selection of CD8SP Vb8.1+ TCR-tg cells could substrates has revealed that the proteome of FAT10 substrates is be observed in the thymus, which led to decreased CD8SP Vb8.1+ very different from that of ubiquitin (35), suggesting that different TCR-tg cells in the spleen (Fig. 5A, 5B). Similarly, negative se- proteins are targeted for Ag presentation by FAT10 and ubiquitin. lection at the DP to CD8SP transition of HY TCR-tg cells was Therefore, FAT10ylation could serve as a backup system for enhanced in male FAT10-deficient mice (Fig. 5C), but not in fe- ubiquitylation to avoid incomplete presentation of self-Ags. Al- male mice (Fig. 4A). Due to a developmental arrest at the DN to ternatively, the mTEC-confined FAT10 could contribute to dif- DP transition, a drastic reduction in DP thymocytes can be ob- ferences in the peptide repertoire presented by mTECs and cTECs, served in male HY TCR-tg mice. Although we found strongly which may be important for efficient T cell selection. Such a role reduced numbers of DP cells in male HY TCR-tg mice, no dif- was previously proposed for the thymoproteasome subunit b5t, ference in DP cells between wild-type and FAT102/2 mice could which is exclusively expressed in cTECs but not mTECs and is be observed. Nevertheless, because CD8SP cells of HY TCR-tg essential for the efficient selection of CD8+ thymocytes (30). cells were reduced in male FAT10-deficient mice (Fig. 5C), but Apart from negative selection of CD8SP T cells (Figs. 4, 5), the not in female mice (Fig. 4A), we conclude that negative selection selection of CD4SP cells was also affected in FAT10-deficient from the DP to CD8SP transition is altered in FAT10-deficient mice. The usage of several Vb-segments of the TCR of CD4SP male HY TCR-tg mice. in the thymus was altered in FAT10-deficient mice (Fig. 3B). How could FAT10 alter negative selection in mTECs? FAT10 Additionally, in the absence of FAT10, increased percentages of was described as a ubiquitin-independent signal for proteasomal OT-II CD4SP in the thymus were observed, suggesting a reduced degradation that directly targets proteins for degradation by the negative selection in these mice. Autophagy in thymic epithelium

Vb8.1/2+ of CD4SP (right panel) by flow cytometry. Shown are the mean percentages 6 SEM (Smarta: n = 21; Smarta FAT102/2: n = 34; Smarta LCMV carrier: n = 7; Smarta FAT102/2 LCMV carrier: n = 1). *p , 0.05, **p , 0.01, ***p , 0.001. 10 FAT10 CODETERMINES THYMIC T CELL SELECTION is an essential process in mediating tolerance of CD4+ cells 4. Raasi, S., G. Schmidtke, and M. Groettrup. 2001. The ubiquitin-like protein FAT10 forms covalent conjugates and induces apoptosis. J. Biol. Chem. 276: (36, 37). It has been demonstrated that the autophagosomal re- 35334–35343. ceptor p62/SQSTM1 becomes covalently mono-FAT10ylated at 5. Schmidtke, G., B. Kalveram, and M. Groettrup. 2009. Degradation of FAT10 by several lysines and that FAT10 colocalizes with p62 in p62 bodies the 26S proteasome is independent of ubiquitylation but relies on NUB1L. FEBS Lett. 583: 591–594. (19). Hence, FAT10 might be involved in targeting Ags for 6. Jin, J., X. Li, S. P. Gygi, and J. W. Harper. 2007. Dual E1 activation systems for autophagosomal degradation and consequently for MHC-II ubiquitin differentially regulate E2 enzyme charging. Nature 447: 1135–1138. presentation on mTECs. To investigate the influence of FAT10 7. Aichem, A., C. Pelzer, S. Lukasiak, B. Kalveram, P. W. Sheppard, N. Rani, + G. Schmidtke, and M. Groettrup. 2010. USE1 is a bispecific conjugating enzyme on negative selection of CD4 cells, we investigated Smarta for ubiquitin and FAT10, which FAT10ylates itself in cis. Nat. Commun. 1: 13. TCR-tg mice neonatally infected with LCMV. Unfortunately, 8. Chiu, Y. H., Q. Sun, and Z. J. Chen. 2007. E1-L2 activates both ubiquitin and the overall survival of LCMV carriers in Smarta mice was FAT10. Mol. Cell 27: 1014–1023. 9. Pelzer, C., I. Kassner, K. Matentzoglu, R. K. Singh, H.-P. Wollscheid, rather low (Table I). Nevertheless, the survival rate in FAT10- M. Scheffner, G. Schmidtke, and M. Groettrup. 2007. UBE1L2, a novel E1 deficient Smarta carrier mice was ∼10 times lower compared with enzyme specific for ubiquitin. J. Biol. Chem. 282: 23010–23014. 10. Ross, M. J., M. S. Wosnitzer, M. D. Ross, B. Granelli, G. L. Gusella, M. Husain, FAT10-proficient Smarta carrier mice (Fig. 5E). Such a survival L. Kaufman, M. Vasievich, V. D. D’Agati, P. D. Wilson, et al. 2006. Role of difference was not observed when the Smarta TCR-tg mice were ubiquitin-like protein FAT10 in epithelial apoptosis in renal disease. J. Am. Soc. not neonatally infected with LCMV. The aforementioned survival Nephrol. 17: 996–1004. 11. Ren, J., Y. Wang, Y. Gao, S. B. K. Mehta, and C. G. L. Lee. 2011. FAT10 difference may hence be attributed to incomplete negative selec- mediates the effect of TNF-a in inducing chromosomal instability. J. Cell Sci. tion of Smarta-tg T cells in FAT10-deficient LCMV carrier mice 124: 3665–3675. (as found in the only survivor we could analyze), resulting in le- 12. Gong, P., A. Canaan, B. Wang, J. Leventhal, A. Snyder, V. Nair, C. D. Cohen, M. Kretzler, V. D’Agati, S. Weissman, and M. J. Ross. 2010. The ubiquitin-like thal autoimmunity. protein FAT10 mediates NF-kappaB activation. J. Am. Soc. Nephrol. 21: 316–326. Downloaded from Given that FAT10 affects negative selection of self-reactive 13. Canaan, A., X. Yu, C. J. Booth, J. Lian, I. Lazar, S. L. Gamfi, K. Castille, T cells in the thymus, it is quite possible that FAT10 deficiency N. Kohya, Y. Nakayama, Y. C. Liu, et al. 2006. FAT10/diubiquitin-like protein- deficient mice exhibit minimal phenotypic differences. Mol. Cell. Biol. 26: or FAT10 mutations result in autoimmunity. So far, we and others 5180–5189. 2 2 have not found evidence for autoimmunity in FAT10 / mice (13). 14. Lukasiak, S., C. Schiller, P. Oehlschlaeger, G. Schmidtke, P. Krause, However, very recently, Cort et al. (38) have shown that the gene D. F. Legler, F. Autschbach, P. Schirmacher, K. Breuhahn, and M. Groettrup. 2008. Proinflammatory cytokines cause FAT10 upregulation in cancers of liver encoding FAT10, called diubiquitin (ubd), lies in a major sus- and colon. Oncogene 27: 6068–6074. http://www.jimmunol.org/ ceptibility locus for autoimmune diabetes triggered in the rat 15. Raasi, S., G. Schmidtke, R. de Giuli, and M. Groettrup. 1999. A ubiquitin-like protein which is synergistically inducible by interferon-g and tumor necrosis strain LEW.1WR1 by infection with the Kilham rat virus. In factor-a. Eur. J. Immunol. 29: 4030–4036. this study, it was shown that the deletion of the ubd gene 16. Schliehe, C., A. Bitzer, M. van den Broek, and M. Groettrup. 2012. Stable an- in LEW.1WR1 rats strongly reduced their susceptibility to virus- tigen is most effective for eliciting CD8+ T-cell responses after DNA vaccination and infection with recombinant vaccinia virus in vivo. J. Virol. 86: 9782–9793. triggered type 1 diabetes. This phenotype of FAT10 deficiency 17. Ebstein, F., A. Lehmann, and P. M. Kloetzel. 2012. The FAT10- and ubiquitin- would be consistent with the increased negative selection we have dependent degradation machineries exhibit common and distinct requirements found in FAT10-deficient male HY-TCR-tg and LCMV-carrying for MHC class I antigen presentation. Cell. Mol. Life Sci. 69: 2443–2454. 18. Basler, M., S. Buerger, and M. Groettrup. 2015. The ubiquitin-like modifier P14-TCR-tg mice (Fig. 5). Accordingly, FAT10 may have en- FAT10 in antigen processing and antimicrobial defense. Mol. Immunol. DOI: hanced the susceptibility to diabetes because it interfered with 10.1016/j.molimm.2015.04.012. by guest on September 23, 2021 complete negative selection of diabetogenic T cells in virus- 19. Aichem, A., B. Kalveram, V. Spinnenhirn, K. Kluge, N. Catone, T. Johansen, and M. Groettrup. 2012. The proteomic analysis of endogenous FAT10 substrates infected LEW.1WR1 rats. identifies p62/SQSTM1 as a substrate of FAT10ylation. J. Cell Sci. 125: 4576– Taken together, in this study, we could assign FAT10 expression 4585. 20. Spinnenhirn, V., H. Farhan, M. Basler, A. Aichem, A. Canaan, and M. Groettrup. in the thymus to terminally differentiated mTECs. We identified 2014. The ubiquitin-like modifier FAT10 decorates autophagy-targeted Salmo- FAT10 as a novel factor involved in T cell repertoire formation nella and contributes to Salmonella resistance in mice. J. Cell Sci. 127: 4883– and negative selection in the thymus, which by this means may 4893. 21. Canaan, A., J. DeFuria, E. Perelman, V. Schultz, M. Seay, D. Tuck, R. A. Flavell, codetermine the susceptibility to autoimmune diseases. M. P. Snyder, M. S. Obin, and S. M. Weissman. 2014. Extended lifespan and reduced adiposity in mice lacking the FAT10 gene. Proc. Natl. Acad. Sci. USA 111: 5313–5318. Acknowledgments 22. Pircher, H., K. Burki,€ R. Lang, H. Hengartner, and R. M. Zinkernagel. 1989. We thank Ulrike Beck for technical assistance, Friederike Berberich- Tolerance induction in double specific T-cell receptor transgenic mice varies Siebelt for stimulating discussions, Stefan Stevanovic for advice and help with antigen. Nature 342: 559–561. 23. Hogquist, K. A., S. C. Jameson, W. R. Heath, J. L. Howard, M. J. Bevan, and with mass spectrometric analysis, Ludger Klein and Martin Aichinger F. R. Carbone. 1994. T cell receptor antagonist peptides induce positive selec- for support and advice on TEC isolation, and Martina Ga¨rtner and Bruno tion. Cell 76: 17–27. Kyewski for human thymus samples. We also thank Allon Canaan and 24. Barnden, M. J., J. Allison, W. R. Heath, and F. R. Carbone. 1998. Defective TCR Sherman Weissman for the contribution of FAT102/2 mice, Ksenija expression in transgenic mice constructed using cDNA-based alpha- and beta- chain under the control of heterologous regulatory elements. Immunol. Jovanovic for kindly providing the Adig mice with permission from Cell Biol. 76: 34–40. Mark S. Anderson, the Bioimaging Center of Konstanz University for 25. Oxenius, A., M. F. Bachmann, R. M. Zinkernagel, and H. Hengartner. 1998. help with microscopy, and the personnel of the animal facility of Konstanz Virus-specific MHC-class II-restricted TCR-transgenic mice: effects on humoral University for animal care. and cellular immune responses after viral infection. Eur. J. Immunol. 28: 390–400. 26. Kisielow, P., H. Bluthmann,€ U. D. Staerz, M. Steinmetz, and H. von Boehmer. 1988. Tolerance in T-cell-receptor transgenic mice involves deletion of non- mature CD4+8+ thymocytes. Nature 333: 742–746. Disclosures 27. Gardner, J. M., J. J. Devoss, R. S. Friedman, D. J. Wong, Y. X. Tan, X. Zhou, The authors have no financial conflicts of interest. K. P. Johannes, M. A. Su, H. Y. Chang, M. F. Krummel, and M. S. Anderson. 2008. Deletional tolerance mediated by extrathymic Aire-expressing cells. Sci- ence 321: 843–847. References 28. Pfaffl, M. W., G. W. Horgan, and L. Dempfle. 2002. Relative expression software tool (REST) for group-wise comparison and statistical analysis of relative ex- 1. Starr, T. K., S. C. Jameson, and K. A. Hogquist. 2003. Positive and negative pression results in real-time PCR. Nucleic Acids Res. 30: e36. selection of T cells. Annu. Rev. Immunol. 21: 139–176. 29. Basler, M., N. Youhnovski, M. Van Den Broek, M. Przybylski, and M. Groettrup. 2. Derbinski, J., and B. Kyewski. 2005. Linking signalling pathways, thymic 2004. Immunoproteasomes down-regulate presentation of a subdominant T cell stroma integrity and autoimmunity. Trends Immunol. 26: 503–506. epitope from lymphocytic choriomeningitis virus. J. Immunol. 173: 3925–3934. 3. Hipp, M. S., B. Kalveram, S. Raasi, M. Groettrup, and G. Schmidtke. 2005. 30. Murata, S., K. Sasaki, T. Kishimoto, S. Niwa, H. Hayashi, Y. Takahama, and FAT10, a ubiquitin-independent signal for proteasomal degradation. Mol. Cell. K. Tanaka. 2007. Regulation of CD8+ T cell development by thymus-specific Biol. 25: 3483–3491. proteasomes. Science 316: 1349–1353. The Journal of Immunology 11

31. Gillard, G. O., and A. G. Farr. 2005. Contrasting models of promiscuous gene 35. Merbl, Y., P. Refour, H. Patel, M. Springer, and M. W. Kirschner. 2013. Profiling of expression by thymic epithelium. J. Exp. Med. 202: 15–19. ubiquitin-like modifications reveals features of mitotic control. Cell 152: 1160–1172. 32. Teh, H. S., P. Kisielow, B. Scott, H. Kishi, Y. Uematsu, H. Bluthmann,€ and 36. Nedjic, J., M. Aichinger, J. Emmerich, N. Mizushima, and L. Klein. 2008. H. von Boehmer. 1988. Thymic major histocompatibility complex antigens and Autophagy in thymic epithelium shapes the T-cell repertoire and is essential for the alpha beta T-cell receptor determine the CD4/CD8 phenotype of T cells. tolerance. Nature 455: 396–400. Nature 335: 229–233. 37. Aichinger, M., C. Wu, J. Nedjic, and L. Klein. 2013. Macroautophagy substrates 33. Anderson, G., and Y. Takahama. 2012. Thymic epithelial cells: working class are loaded onto MHC class II of medullary thymic epithelial cells for central heroes for T cell development and repertoire selection. Trends Immunol. 33: tolerance. J. Exp. Med. 210: 287–300. 256–263. 38. Cort, L., M. Habib, R. A. Eberwine, M. J. Hessner, J. P. Mordes, and 34. Schmidtke, G., A. Aichem, and M. Groettrup. 2014. FAT10ylation as a signal for E. P. Blankenhorn. 2014. Diubiquitin (Ubd) is a susceptibility gene for virus- proteasomal degradation. Biochim. Biophys. Acta 1843: 97–102. triggered autoimmune diabetes in rats. Genes Immun. 15: 168–175. Downloaded from http://www.jimmunol.org/ by guest on September 23, 2021 The Ubiquitin-like Modifier FAT10 is Selectively Expressed in Medullary Thymic Epithelial Cells and Modifies T Cell Selection

Stefanie Buerger, Valerie L. Herrmann, Sarah Mundt, Nico Trautwein, Marcus Groettrup, and Michael Basler

Supplemental information

Supplemental Figure 1. Normal structural organization of the thymus of FAT10-deficient mice (A) TECs from C57BL/6 wild type and FAT10-/- mice were analyzed by flow cytometry according to the surface markers CD45-EpCAM+Ly-51- (mTECs) and CD45-EpCAM+Ly-51+ (cTECs). Numbers indicate percentages of the respective populations. (B) 5 µm cryosections of C57BL/6 wild type and FAT10-/- mice were co-stained with mouse anti-Keratin 18 (cortex) and rabbit anti-Keratin 5 (medulla) followed by secondary antibodies goat anti-mouse Alexa488 and anti-rabbit Cy3 and analyzed by confocal microscopy. One representative image out of three independent experiments is displayed. Bar, 50 μm. (C) 5 µm cryosections of C57BL/6 wild type and FAT10-/- mice were co-stained with mouse anti-Keratin 8 (cortex) and rabbit anti-Keratin 5 (medulla) followed by secondary antibodies goat anti-mouse Alexa488 and anti- rabbit Cy3 and analyzed by confocal microscopy. One representative image out of three independent experiments is displayed. Bar, 50 μm.

Supplemental Figure 2. Human mTECs co-express Aire and FAT10 (A,B) 5 μm cryosections of human pediatric patients were co-stained with mouse anti-human FAT10 (clone 4F1) and rabbit-anti-Keratin 5 and rat-anti-Aire antibodies followed by secondary antibodies goat anti- mouse-Alexa488 and donkey anti-rabbit Cy3 and goat anti-rat Alexa 633 and analyzed by confocal microscopy. Patient numbers are indicated (left side). One representative image out of three independent experiments is displayed. (A) Images were taken with a 63x multi-immersion objective. Bar, 50 μm. (B) Images were taken with a 25x multi-immersion objective to give a broad overview of the thymic architecture. Bar, 50 µm.

Supplemental Figure 3. Thymocyte and regulatory T cell (Treg) distribution in C57BL/6 and FAT10-/- mice. (A) Thymocytes derived from C57BL/6 and FAT10-/- mice were stained for CD4 and CD8 and analyzed by flow cytometry. Representative flow cytometry plots are shown on the left. Right graphs show the mean percentage ± s.e.m of DN, DP, CD4SP, or CD8SP cells of thymocytes derived from C57BL/6 or FAT10-/- (n=5 per group). (B) Absolute cell numbers per thymus of DN, DP, CD4SP, or CD8SP derived from C57BL/6 or FAT10-/- mice. Values show mean ± s.e.m. (n=6 per group). (C) CD4SP thymocytes derived from C57BL/6 or FAT10-/- mice were analyzed for Foxp3 expression by intracellular staining. Representative flow cytometry plots are shown on the left. Right graphs show the mean percentage ± s.e.m of Foxp3+ of CD4SP cells derived from C57BL/6 or FAT10-/- mice (n=5 per group). n.s.: not significant. (D) Vβ expression on DP thymocytes. Flow cytometric analysis of indicated Vβ variable segments of TCRs from DP thymocytes derived from C57BL/6 or FAT10-/- mice. The graphs shows the mean ± s.e.m of two independent experiments (n=10). (E) Absolute number of thymocytes of TCRtg. Absolute number of FAT10+/+ or FAT10-/- thymocytes derived from OT-I, P14, HY male (m), HY female (f), OT-II, and Smarta were determined by flow cytometry. Shown are the mean percentages ± s.e.m of number of thymocytes per thymus. (OT-I FAT10+/+: n=10; OT-I FAT10-/-: n=10; P14 FAT10+/+: n=10; P14 FAT10-/-: n=10; HY FAT10+/+ m: n=6; HY FAT10-/- m: n=8; HY FAT10+/+ f: n=7; HY FAT10-/- f: n=7; OT-II FAT10+/+: n=5; OT-II FAT10-/-: n=5; Smarta FAT10+/+: n=6; Smarta FAT10-/-: n=6). Supplemental Table 1; Related to Figure 6. Identification of peptides eluted from HLA molecules of wild type and stably FAT10 - transfected HEK293 cells.

Peptides unique to wild type HEK293 cells

Accession Peptide Modifications MH+ [Da] Protein number

O14777 GLNEEIARV 1000,542 Kinetochore protein NDC80 homolog O14925 GTMTGMLYK 1001,480 Mitochondrial import inner membrane translocase subunit Tim23 O60547 FLLEKGYEV 1097,588 GDP-mannose 4,6 dehydratase O95997 KTKGPLKQK 1027,662 Securin P05067 LLAAWTARA 972,562 Amyloid beta A4 protein P08670 NLAEDIMRL 1074,561 Vimentin P11021 KLSLVAAML 945,580 78 kDa glucose-regulated protein P12004 KLMDLDVEQL 1203,629 Proliferating cell nuclear antigen P17844 FRTGNPTGTY 1113,532 Probable ATP-dependent RNA helicase DDX5 P18859 PKFEVIEKPQA 1285,715 ATP synthase-coupling factor 6, mitochondrial P19338 VKLAKAGKN 928,593 Nucleolin P20700 RLAVYIDKV 1076,646 Lamin-B1 P29692 VRIASLEVE 1015,578 Elongation factor 1-delta P35251 SPKASSKL 817,478 Replication factor C subunit 1 P40937 KMADIEYRL 1138,592 Replication factor C subunit 5 P40937 ALHDILTEI 1024,567 Replication factor C subunit 5 P43246 RLYQGINQL 1104,616 DNA mismatch repair protein Msh2 P60468 KVGPVPVLV 907,597 Protein transport protein Sec61 subunit beta P61313 ATYGKPVHH 1009,521 60S ribosomal protein L15 P61421 RLYPEGLAQL 1159,647 V-type proton ATPase subunit d 1 P62424 KVAPAPAVVK 979,629 60S ribosomal protein L7a P63241 AVAIKAmAK M7(Oxidation) 918,544 Eukaryotic translation initiation factor 5A-1 P69849 KIAPNTPQL 981,572 Nodal modulator 3 Q00839 ASYGVSKGK 896,483 Heterogeneous nuclear ribonucleoprotein U Q12769 AVYDRPGASPK 1160,605 Nuclear pore complex protein Nup160 Q15011 SLLPEGPPAI 993,561 Homocysteine-responsive endoplasmic reticulum-resident ubiquitin- like domain member 1 protein Q5SRE5 NPAKVWTDL 1043,552 Nucleoporin NUP188 homolog Q5VYK3 NLAEKPKTV 999,583 Proteasome-associated protein ECM29 homolog Q6NUQ1 RTNWPNTGK 1073,549 RAD50-interacting protein 1 Q7KZF4 SPAFSTRVL 977,541 Staphylococcal nuclease domain-containing protein 1 Q8NI27 NLIDLDDLYV 1192,610 THO complex subunit 2 Q8TE76 GLNNKTIGY 979,521 MORC family CW-type zinc finger protein 4 Q92688 SLINVGLISV 1014,619 Acidic leucine-rich nuclear phosphoprotein 32 family member B Q96C01 RVQEAVESMVK 1275,673 Protein FAM136A Q96HC4 VPRQPTVTSV 1083,616 PDZ and LIM domain protein 5 Q99504 APAPAAQRL 894,516 Eyes absent homolog 3 Q99873 KLDHVVTII 1037,635 Protein arginine N-methyltransferase 1 Q9NV31 ALLDKLYAL 1019,613 U3 small nucleolar ribonucleoprotein protein IMP3 Q9UQB9 KIADFGWSV 1022,531 Aurora kinase C Q9Y520 VPPPPHRPL 1009,594 Protein PRRC2C Q9Y6A5 KLVEFDFLGA 1138,614 Transforming acidic coiled-coil-containing protein 3 Q9Y6W5 PPPAPPPPP 866,477 Wiskott-Aldrich syndrome protein family member 2 Peptides unique to HEKFAT10 cells

Accession Peptide Modifications MH+ [Da] Protein name number

P05386 AKALANVN 800,463 60S acidic ribosomal protein P1 P11279 APGSARRPL 924,536 Lysosome-associated membrane glycoprotein 1 P62158 FVQmmTAK M4(Oxidation); 987,463 Calmodulin M5(Oxidation) P62854 APRPPPKPm M9(Oxidation) 1006,549 40S ribosomal protein S26 Q15366 RLSSETGGMGSS 1168,526 Poly(rC)-binding protein 2 Q96BR5 KPGKKSIAA 899,567 Cytochrome c oxidase assembly factor 7 Q9C0D2 SPAIGRTSI 901,510 Centrosomal protein KIAA1731 Peptides shared by wild type HEK293 cells and HEKFAT10 cells

Accession Peptide Modifications MH+ [Da] Protein name number

A5YKK6 APFLRNVEL 1058,599 CCR4-NOT transcription complex subunit 1 A6NFI3 APAPKPEAA 851,462 Zinc finger protein 316 A6NKF1 KPRPPPSQL 1019,599 SAC3 domain-containing protein 1 A8MPP1 SPTGTGKSL 847,452 Putative ATP-dependent RNA helicase DDX11-like protein 8 O00186 ALGTDAEGQKV 1088,557 Syntaxin-binding protein 3 O00232 RPKDPNNLL 1066,600 26S proteasome non-ATPase regulatory subunit 12 O00303 ALNEKLVNL 1013,599 Eukaryotic translation initiation factor 3 subunit F O00487 AAMLDTVVFK 1094,591 26S proteasome non-ATPase regulatory subunit 14 O00487 FVDDYTVRV 1113,558 26S proteasome non-ATPase regulatory subunit 14 O14497 SPRGGTPGSGAA 1014,496 AT-rich interactive domain-containing protein 1A O14604 KLKKTETQEK 1232,721 Thymosin beta-4, Y-chromosomal O14980 VLIDYQRNV 1119,616 Exportin-1 O15042 RPKKPGQSF 1044,595 U2 snRNP-associated SURP motif-containing protein O15121 FPNIPGKSL 972,551 Sphingolipid delta(4)-desaturase DES1 O15131 VMDSKIVQV 1018,560 Importin subunit alpha-6 O15391 KTLEGEFSV 1009,520 Transcription factor YY2 O15539 GLASFKSFLK 1097,635 Regulator of G-protein signaling 5 O43324 SLLEKSLGL 959,577 Eukaryotic translation elongation factor 1 epsilon-1 O60341 NRLLEATSY 1066,552 Lysine-specific histone demethylase 1A O60508 SPSSKPSL 802,431 Pre-mRNA-processing factor 17 O60925 KVIDTQQKV 1058,620 Prefoldin subunit 1 O60925 KLADIQIEQL 1170,673 Prefoldin subunit 1 O75643 APTGSGKTI 831,457 U5 small nuclear ribonucleoprotein 200 kDa helicase O75807 VPRGQGSQF 975,501 Protein phosphatase 1 regulatory subunit 15A O76080 SASVQRADTSL 1134,574 AN1-type zinc finger protein 5 O95147 TVADKIHSV 969,536 Dual specificity protein phosphatase 14 O95817 LPSSGRSSL 903,489 BAG family molecular chaperone regulator 3 O95848 SPYLRPLTL 1059,618 Uridine diphosphate glucose pyrophosphatase P04818 EPRPPHGEL 1031,526 Thymidylate synthase P04844 RLDELGGVYL 1134,615 Dolichyl-diphosphooligosaccharide--protein glycosyltransferase subunit 2 P07203 LLIENVASL 971,577 Glutathione peroxidase 1 P07225 FLSKQQASQV 1135,610 Vitamin K-dependent protein S P07686 ALVVQVAEA 899,519 Beta-hexosaminidase subunit beta P07910 SNVTNKTD 878,421 Heterogeneous nuclear ribonucleoproteins C1/C2 P08579 NQFPGFKEV 1065,536 U2 small nuclear ribonucleoprotein B'' P08670 RPSTSRSL 903,500 Vimentin P09874 APRGKSGAAL 927,537 Poly [ADP-ribose] polymerase 1 P09958 AVRIPGGPA 837,494 Furin P0C6E5 APKRPPSGF 956,531 Putative high mobility group protein B3-like protein P0CW22 QVTQPTVGmN M9(Oxidation) 1090,519 40S ribosomal protein S17-like P10398 KIGDFGLATV 1020,572 Serine/threonine-protein kinase A-Raf P11217 NLAENISRV 1015,553 Glycogen phosphorylase, muscle form P11926 ILDQKINEV 1071,604 Ornithine decarboxylase P13284 LLLDVPTAAVQA 1210,704 Gamma-interferon-inducible lysosomal thiol reductase P13378 HPPPPPPPP 932,498 Homeobox protein Hox-D8 P13378 PHPPPPPPP 932,498 Homeobox protein Hox-D8 P13639 ILTDITKGV 959,577 Elongation factor 2 P13639 SPNKHNRL 965,527 Elongation factor 2 P13639 SPVVRVAV 826,514 Elongation factor 2 P13796 ALPEDLVEV 984,525 Plastin-2 P13987 SLSEKTVLL 989,587 CD59 glycoprotein P13995 AVIDVGINRV 1055,620 Bifunctional methylenetetrahydrofolate dehydrogenase/cyclohydrolase, mitochondrial P16402 KPRKPAGAA 895,547 Histone H1.3 P16403 AASKERSGVSL 1104,601 Histone H1.2 P16403 ASKERSGVSL 1104,601 Histone H1.2 P17844 YLLPAIVHI 1038,634 Probable ATP-dependent RNA helicase DDX5 P18085 KLGEIVTTI 973,592 ADP-ribosylation factor 4 P18124 VPAEPKLAF 971,555 60S ribosomal protein L7 P23588 LPKSPPYTAF 1120,603 Eukaryotic translation initiation factor 4B P23921 RPAANPIQF 1013,552 Ribonucleoside-diphosphate reductase large subunit P26358 GLIEKNIEL 1028,598 DNA (cytosine-5)-methyltransferase 1 P26373 APSRNGmVL M7(Oxidation) 960,493 60S ribosomal protein L13 P26373 APRPASGPIRP 1118,643 60S ribosomal protein L13 P26583 APKRPPSAF 970,547 High mobility group protein B2 P26599 GVYGDVQRV 992,516 Polypyrimidine tract-binding protein 1 P27708 LATVLGRF 876,530 CAD protein P27708 GLADKVYFL 1025,567 CAD protein P27816 RPEEGRPVV 1038,569 Microtubule-associated protein 4 P29558 SPAAAQKAV 842,473 RNA-binding motif, single-stranded-interacting protein 1 P33176 ALSEELVQL 1001,551 Kinesin-1 heavy chain P33316 RPAEVGGmQL M8(Oxidation) 1073,540 Deoxyuridine 5'-triphosphate nucleotidohydrolase, mitochondrial P33552 KYFDEHYEY 1293,543 Cyclin-dependent kinases regulatory subunit 2 P35637 DYTQQATQSYG 1261,533 RNA-binding protein FUS P35659 NRPGTVSSL 930,500 Protein DEK P38606 SLAETDKITL 1090,600 V-type proton ATPase catalytic subunit A P40939 SPNSKVNTL 959,515 Trifunctional enzyme subunit alpha, mitochondrial P41091 ALPEIFTEL 1032,561 Eukaryotic translation initiation factor 2 subunit 3 P41221 AmSSKFFLV M2(Oxidation) 1045,538 Protein Wnt-5a P43686 RPQTGLSFL 1018,568 26S protease regulatory subunit 6B P46013 RPKSGGSGHAV 1052,559 Antigen KI-67 P46060 RVIGTLEEV 1015,578 Ran GTPase-activating protein 1 P46777 KIYEGQVEV 1064,562 60S ribosomal protein L5 P46977 KLNPQQFEV 1102,589 Dolichyl-diphosphooligosaccharide--protein glycosyltransferase subunit STT3A P48444 RPSGPSKAL 912,526 Coatomer subunit delta P49327 IPRDPSQQEL 1182,611 Fatty acid synthase P49327 SPAPTHNSL 923,458 Fatty acid synthase P49720 GLATDVQTV 903,477 Proteasome subunit beta type-3 P49755 LLGPRLVLA 951,634 Transmembrane emp24 domain-containing protein 10 P49815 SLLDIIEKV 1029,619 Tuberin P50914 VPAKKITAA 898,572 60S ribosomal protein L14 P51531 RPSGPGPEL 909,479 Probable global transcription activator SNF2L2 P51570 SPRKDGLVSL 1071,615 Galactokinase P51572 KLDVGNAEV 944,504 B-cell receptor-associated protein 31 P51665 RIGKVGNQK 999,605 26S proteasome non-ATPase regulatory subunit 7 P52788 SPYQNIKIL 1075,613 Spermine synthase P53350 KIGDFGLATK 1049,599 Serine/threonine-protein kinase PLK1 P53675 NPASKVIAL 912,551 Clathrin heavy chain 2 P53814 SPGGPRAAV 811,442 Smoothelin P54289 LENDIKVEI 1072,588 Voltage-dependent calcium channel subunit alpha-2/delta-1 P54819 FLLDGFPRTV 1164,641 Adenylate kinase 2 P60709 APEEHPVL 891,457 Actin, cytoplasmic 1 P61247 SLADLQNDEV 1103,521 40S ribosomal protein S3a P61978 VPDSSGPERIL 1169,616 Heterogeneous nuclear ribonucleoprotein K P62136 SIIGRLLEV 999,619 Serine/threonine-protein phosphatase PP1-alpha catalytic subunit P62241 RIIDVVYNA 1062,594 40S ribosomal protein S8 P62249 KLLEPVLLL 1037,696 40S ribosomal protein S16 P62424 KVAPAPAVV 851,535 60S ribosomal protein L7a P62491 ALDSTNVEA 919,437 Ras-related protein Rab-11A P62750 APAPPKAEA 851,462 60S ribosomal protein L23a P62906 NmVAKVDEV M2(Oxidation) 1020,503 60S ribosomal protein L10a P62906 TLYEAVREV 1079,572 60S ribosomal protein L10a P62917 APAGRKVGL 868,536 60S ribosomal protein L8 P62917 RVKLPSGSKK 1099,695 60S ribosomal protein L8 P63092 ALWEDEGVRA 1145,558 Guanine nucleotide-binding protein G(s) subunit alpha isoforms short P63220 MNVAEVDKV 1004,508 40S ribosomal protein S21 P63272 LPQGIVREL 1024,615 Transcription elongation factor SPT4 P69849 SVNGKVLSK 931,557 Nodal modulator 3 P78347 KQVEEIFNL 1119,604 General transcription factor II-I P78543 TLWVDPYEV 1121,551 Protein BTG2 P86790 SPSKPAKSL 914,530 Vacuolar fusion protein CCZ1 homolog B Q01581 RPTGGVGAV 813,458 Hydroxymethylglutaryl-CoA synthase, cytoplasmic Q01844 FRQDHPSSm M9(Oxidation) 1120,484 RNA-binding protein EWS Q04637 AARPATSTL 887,494 Eukaryotic translation initiation factor 4 gamma 1 Q05048 LPDERTISL 1043,572 Cleavage stimulation factor subunit 1 Q07021 APASPFRQL 986,541 Complement component 1 Q subcomponent-binding protein, mitochondrial Q07820 RPPPIGAEV 935,531 Induced myeloid leukemia cell differentiation protein Mcl-1 Q08211 KVFDPVPVGV 1056,609 ATP-dependent RNA helicase A Q10570 FLPSYIIDV 1066,581 Cleavage and polyadenylation specificity factor subunit 1 Q12789 NPKESSSSL 948,463 General transcription factor 3C polypeptide 1 Q13111 SPRSPSTTYL 1108,563 Chromatin assembly factor 1 subunit A Q13428 APAGTRSQV 886,474 Treacle protein Q13868 FLQEGDLISA 1092,557 Exosome complex component RRP4 Q14145 GVIDGHIYAV 1043,552 Kelch-like ECH-associated protein 1 Q14257 GPRTAALGLL 968,587 Reticulocalbin-2 Q14669 SPRLPVGGF 929,520 E3 ubiquitin-protein ligase TRIP12 Q14807 SPNAEIHIL 993,536 Kinesin-like protein KIF22 Q14C87 RPKQEAAI 912,525 Transmembrane protein 132D Q15021 SLAGDVALQQL 1114,610 Condensin complex subunit 1 Q15041 YLDPSVLSGV 1049,551 ADP-ribosylation factor-like protein 6-interacting protein 1 Q15084 VELDDLGKDEL 1245,622 Protein disulfide-isomerase A6 Q15366 SLAQYLINV 1020,572 Poly(rC)-binding protein 2 Q15370 RPQAPATVGL 1009,579 Transcription elongation factor B polypeptide 2 Q15393 SLLGDDALVQV 1129,610 Splicing factor 3B subunit 3 Q15437 SLLPPDALVGL 1094,645 Protein transport protein Sec23B Q15717 NPNQNKNVAL 1111,585 ELAV-like protein 1 Q16531 RPKGESKDL 1029,569 DNA damage-binding protein 1 Q16531 RPSASTQAL 930,500 DNA damage-binding protein 1 Q17RS7 NVRPPNTAL 981,547 Flap endonuclease GEN homolog 1 Q29RF7 AIDPHLLLSV 1077,630 Sister chromatid cohesion protein PDS5 homolog A Q4VCS5 NPSENRSLL 1029,532 Angiomotin Q53EZ4 SPKSPTAAL 871,488 Centrosomal protein of 55 kDa Q58FF7 ILDKKVEKV 1071,676 Putative heat shock protein HSP 90-beta-3 Q5EBM0 LLKPDLILL 1037,697 UMP-CMP kinase 2, mitochondrial Q5H9R7 APRPPSSSP 895,463 Serine/threonine-protein phosphatase 6 regulatory subunit 3 Q5JSZ5 RPAGGNGSGL 885,454 Protein PRRC2B Q5SQI0 ALFPERITV 1045,604 Alpha-tubulin N-acetyltransferase 1 Q5T1J5 APRQPGLmA M8(Oxidation) 956,498 Putative coiled-coil-helix-coiled-coil-helix domain-containing protein CHCHD2P9, mitochondrial Q68D10 GLGPPGRSV 839,473 Protein SPT2 homolog Q6NUT3 APSPRPLSL 937,546 Major facilitator superfamily domain-containing protein 12 Q6NW40 TPRGGSDL 802,405 RGM domain family member B Q6NZI2 SLLDKIIGA 929,566 Polymerase I and transcript release factor Q6P2C8 RPKAQPTTL 1011,594 Mediator of RNA polymerase II transcription subunit 27 Q6P995 VLLKARLVPA 1079,730 Protein FAM171B Q71SY5 GPARAGSVV 813,458 Mediator of RNA polymerase II transcription subunit 25 Q86U06 KLAEGAGIQL 999,583 Probable RNA-binding protein 23 Q86WA8 SLQSTILGV 917,530 Lon protease homolog 2, peroxisomal Q86X10 NLAEKLIGV 956,576 Ral GTPase-activating protein subunit beta Q86Y91 IARLPSSTL 957,572 Kinesin-like protein KIF18B Q8N4T4 AANSERSQTTL 1177,581 Rho guanine nucleotide exchange factor 39 Q8NCT1 RPKVPDQSV 1025,573 Arrestin domain-containing protein 4 Q8ND30 APNLRGSGV 870,479 Liprin-beta-2 Q8TEL6 TLIEDILGV 972,561 Short transient receptor potential channel 4-associated protein Q8TF50 APRLPITGL 937,583 Zinc finger protein 526 Q8WUU4 SPRASGSGL 831,432 Zinc finger protein 296 Q8WVX3 RPQSGANGL 899,469 Uncharacterized protein C4orf3 Q8WYA0 ALASVIKEL 943,582 Intraflagellar transport protein 81 homolog Q92499 LHLGYLPNQL 1167,652 ATP-dependent RNA helicase DDX1 Q92503 QLIDKVWQL 1142,657 SEC14-like protein 1 Q92522 SLAKIYTEA 995,541 Histone H1x Q92621 ALLDRIVSV 985,604 Nuclear pore complex protein Nup205 Q92624 AYSSYVHQY 1117,495 Amyloid protein-binding protein 2 Q92743 IPRAALLPLL 1076,719 Serine protease HTRA1 Q92830 NPRIPYTEL 1102,589 Histone acetyltransferase KAT2A Q92888 SPGPSRPGL 867,468 Rho guanine nucleotide exchange factor 1 Q93063 SLFRVITEV 1063,614 Exostosin-2 Q96B01 SVKSPNQSL 959,516 RAD51-associated protein 1 Q96MM7 AASRPGSVAA 886,474 Heparan-sulfate 6-O-sulfotransferase 2 Q96NU1 SPYGGGHAL 858,410 Sterile alpha motif domain-containing protein 11 Q96PK6 YRAQPSVSL 1020,547 RNA-binding protein 14 Q96QV1 FILEKEGYV 1097,587 Hedgehog-interacting protein Q96S82 GPRPITQSEL 1097,594 Ubiquitin-like protein 7 Q96T58 SPRPSGPGPSSF 1172,570 Msx2-interacting protein Q99567 ILDPHVVLL 1018,629 Nuclear pore complex protein Nup88 Q99576 SLLGGDVVSV 945,525 TSC22 domain family protein 3 Q99613 SLDQPTQTV 988,494 Eukaryotic translation initiation factor 3 subunit C Q99707 KPRVPPATAF 1083,631 Methionine synthase Q99856 SPKLPVSSL 927,550 AT-rich interactive domain-containing protein 3A Q9BPX6 RPTTGNTL 859,463 Calcium uptake protein 1, mitochondrial Q9BRQ6 AIQDKLFQV 1061,599 Coiled-coil-helix-coiled-coil-helix domain-containing protein 6, mitochondrial Q9BSJ5 AAAAGRKTL 858,515 Uncharacterized protein C17orf80 Q9BVI0 RPKETDHKSL 1210,654 PHD finger protein 20 Q9BVK6 RPRPGTGLGRVm M12(Oxidation) 1312,726 Transmembrane emp24 domain-containing protein 9 Q9BWF3 VRTPYTmSY M7(Oxidation) 1133,530 RNA-binding protein 4 Q9BWN1 SPRPPAEGF 957,479 Proline-rich protein 14 Q9BX63 SPTGSGKSL 833,436 Fanconi anemia group J protein Q9GZM3 LLFEGEKITI 1162,672 DNA-directed RNA polymerase II subunit RPB11-b1 Q9GZR7 SPAKNPSSL 900,478 ATP-dependent RNA helicase DDX24 Q9H4L4 TPSRKGLVL 970,604 Sentrin-specific protease 3 Q9H6E5 RPSGLHGDVSL 1137,601 Speckle targeted PIP5K1A-regulated poly(A) polymerase Q9H7N4; PSPPPPPPP 882,472 Splicing factor, arginine/serine-rich 19; Proline-rich protein 12 Q9ULL5 Q9HB71 VKTDTVLIL 1001,623 Calcyclin-binding protein Q9HDC5 SPAGTRGGF 849,421 Junctophilin-1 Q9NR09 SPRVPNSSV 942,500 Baculoviral IAP repeat-containing protein 6 Q9NRK6 RLYDPASGTISL 1292,685 ATP-binding cassette sub-family B member 10, mitochondrial Q9NRL3 APRAPPGPAGL 1003,567 Striatin-4 Q9NRY4 GLSTEGIYRV 1094,583 Rho GTPase-activating protein 35 Q9NTK1 TIRETTEEm M9(Oxidation) 1125,508 Protein DEPP Q9NVQ4 ALSDGVHKI 939,525 Fas apoptotic inhibitory molecule 1 Q9NX24 EEVQSLPLPL 1124,619 H/ACA ribonucleoprotein complex subunit 2 Q9NZZ3 KPKAPPPSL 934,572 Charged multivesicular body protein 5 Q9P270 SPRGFPLGL 943,536 SLAIN motif-containing protein 2 Q9P2G9 TPRGGVGSV 829,453 Kelch-like protein 8 Q9UBF2 RLPDDDPTAV 1098,543 Coatomer subunit gamma-2 Q9UBU8 APPEYHRKAV 1167,627 Mortality factor 4-like protein 1 Q9UBW8 ALATLIHQV 965,577 COP9 signalosome complex subunit 7a Q9UII2 SIREAGGAF 907,463 ATPase inhibitor, mitochondrial Q9UKL0 APNNGQNKSL 1042,527 REST corepressor 1 Q9ULA0 PSLSHNLLVD 1094,584 Aspartyl aminopeptidase Q9ULL0 PRQLSQAL 912,526 Uncharacterized protein KIAA1210 Q9ULV3 LLGPPPVGV 848,523 Cip1-interacting zinc finger protein Q9UQ16 TLIDLPGITKV 1169,714 Dynamin-3 Q9UQE7 KLDQDLNEV 1073,547 Structural maintenance of protein 3 Q9Y3U8 KVSKDKRAL 1044,652 60S ribosomal protein L36 Q9Y4A5; PAPPPPPPP 866,477 Transformation/transcription domain-associated protein; Q96Q04 Serine/threonine-protein kinase LMTK3 Q9Y4C2 KLGSVPVTV 899,556 Protein FAM115A Q9Y508 APGVRAVEL 911,531 E3 ubiquitin-protein ligase RNF114 Q9Y617 KVQAGNSSL 903,489 Phosphoserine aminotransferase Q9Y678 AIVDKVPSV 927,551 Coatomer subunit gamma-1 Q9Y6G9 SPRVPGGSP 853,453 Cytoplasmic dynein 1 light intermediate chain 1 Q9Y6P5 SLAELVHAV 938,530 Sestrin-1 Q9Y6W5 APPPPPPPP 866,477 Wiskott-Aldrich syndrome protein family member 2