Functional Role for IκBNS in T Cell Cytokine Regulation As Revealed by Targeted Gene Disruption
This information is current as Maki Touma, Valeria Antonini, Manoj Kumar, Stephanie L. of October 2, 2021. Osborn, April M. Bobenchik, Derin B. Keskin, John E. Connolly, Michael J. Grusby, Ellis L. Reinherz and Linda K. Clayton J Immunol 2007; 179:1681-1692; ; doi: 10.4049/jimmunol.179.3.1681 http://www.jimmunol.org/content/179/3/1681 Downloaded from
References This article cites 61 articles, 28 of which you can access for free at:
http://www.jimmunol.org/content/179/3/1681.full#ref-list-1 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 October 2, 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 © 2007 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology
Functional Role for IBNS in T Cell Cytokine Regulation As Revealed by Targeted Gene Disruption1
Maki Touma,*† Valeria Antonini,*† Manoj Kumar,2*† Stephanie L. Osborn,3* April M. Bobenchik,4* Derin B. Keskin,*† John E. Connolly,‡ Michael J. Grusby,§ Ellis L. Reinherz,*† and Linda K. Clayton5*†
Triggering of the TCR by cognate peptide/MHC ligands induces expression of IBNS, a member of the IB family of NF-B inhibitors whose expression is associated with apoptosis of immature thymocytes. To understand the role of IBNS in TCR triggering, we created a targeted disruption of the IBNS gene. Surprisingly, mice lacking IBNS show normal thymic progression but both thymocytes and T cells manifest reduced TCR-stimulated proliferation. Moreover, IBNS knockout thymocytes and T cells produce significantly less IL-2 and IFN-␥ than wild-type cells. Transfection analysis demonstrates that I BNS and c-Rel individually increase IL-2 promoter activity. The effect of I BNS on the IL-2 promoter, unlike c-Rel, Downloaded from is dependent on the NF-B rather than the CD28RE site; mutation of the NF-B site extinguishes the induction of tran- scription by IBNS in transfectants and prevents association of IBNS with IL-2 promoter DNA. Microarray analyses confirm the reduction in IL-2 production and some IFN-␥-linked transcripts in IBNS knockout T cells. Collectively, our findings demonstrate that IBNS regulates production of IL-2 and other cytokines induced via “strong” TCR ligation. The Journal of Immunology, 2007, 179: 1681–1692. http://www.jimmunol.org/ ctivation of the TCR upon Ag recognition initiates a nals emanating from the same molecule lead to two such opposite complex signaling cascade the outcome of which is af- outcomes remains a major question in immunology. A fected by a variety of factors. These include the devel- pMHC engagement of the TCR induces multiple posttransla- opmental stage of the T cell, costimulation by other cell surface tional biochemical modifications and new gene transcription (3) receptors on both the T cell and APC and the strength of the TCR- and the ubiquitous transcription factor, NF-B, has been impli- peptide/MHC (pMHC)6 interaction. In double-positive (DP) thy- cated in both survival and death of thymocytes (4–6). Previously, mocytes, a relatively strong signal induces negative selection and our analysis of genes induced in DP thymocytes undergoing cog- results in death of the thymocyte via apoptosis while a weak TCR nate peptide-induced apoptosis identified a novel member of the signal, by comparison, induces positive selection leading to DP IB family of NF-B inhibitors, IBNS (7). Induction of IBNS by guest on October 2, 2021 thymocyte differentiation and survival (reviewed in Refs. 1 and 2). expression in the thymus correlated with TCR signal strength and The same pMHC ligand that induces apoptosis in DP thymocytes relatively avid pMHC ligands triggered expression of IBNS activates peripheral T cells to proliferate. Elucidation of how sig- whereas weaker ligands did not. In fetal thymic organ cultures (FTOC) retrovirally infected to overexpress IBNS, development of double-negative (DN) to DP thymocytes was reduced and ad- *Laboratory of Immunobiology, Department of Medical Oncology, Dana Farber Can- dition of anti-CD3 mAb had a much more dramatic apoptotic cer Institute, and †Department of Medicine, Harvard Medical School, Boston, MA effect on IBNS-expressing than control DP thymocytes. Thus, 02115; ‡Baylor Institute for Immunology Research, Dallas, TX 75204; and §Depart- ment of Immunology and Infectious Diseases, Harvard School of Public Health, Bos- expression of I BNS alone is not sufficient for DP thymocyte ton, MA 02115 death, rather additional signaling via the TCR complex is required Received for publication April 12, 2007. Accepted for publication May 18, 2007. to regulate the response outcome. The costs of publication of this article were defrayed in part by the payment of page The NF-B protein family plays a key role in the overall reg- charges. This article must therefore be hereby marked advertisement in accordance ulation of the immune system. Activation through the TCR is de- with 18 U.S.C. Section 1734 solely to indicate this fact. pendent on NF-B for proliferation and cytokine production (8– 1 This work was supported by National Institutes of Health Grants AI19807 (to 10). The pathway leading from the TCR to NF-B activation E.L.R.) and AI51779 (to L.K.C.). involves multiple components culminating in the ubiquitination 2 Current address: Beth Israel Deaconess Medical Center Genomic Center, Depart- ment of Medicine, Harvard Medical School, Boston, MA 02115. and degradation of IB inhibitors, thereby releasing NF-Btoen- 3 Current address: Department of Molecular and Cell Biology, Division of Immunol- ter the nucleus and direct gene transcription. This process is com- ogy and Cancer Research Laboratory, University of California, Berkeley, CA 94720. plicated by the presence of five NF-B family members, each ca- 4 Current address: Center for Vascular Biology, University of Connecticut Health pable of forming multiple hetero- and homodimers (11). The Center, Farmington, CT 06030. complexity is amplified by the fact that while specific NF-B- 5 Address correspondence and reprint requests to Dr. Linda K. Clayton, Dana Farber binding DNA sites are preferentially bound by particular dimers, Cancer Institute, 77 Avenue Louis Pasteur, Boston, MA 02115. E-mail address: [email protected] multiple varieties of dimers can also bind with comparable affin- 6 Abbreviations used in this paper: pMHC, peptide-MHC complex; DP, double pos- ities to the same DNA sequence (12–14). Functions unique to one itive; DN, double negative; LN, lymph node; WT, wild type; KO, knockout; BM, NF-B dimer may result from the range of DNA sequences bound bone marrow; FC, fold change; rh, recombinant human; SAP, SLAM-associated at high affinity by that protein dimer (15). Thus, gene regulation at protein. a specific NF-B site is dependent on the identities and amounts of Copyright © 2007 by The American Association of Immunologists, Inc. 0022-1767/07/$2.00 dimers expressed in the cell; this can also vary over time as the
www.jimmunol.org 1682 IBNS REGULATION OF T CELL CYTOKINES accumulation or destruction of some dimers results in NF-B VSV8 peptide for 2 h. After 24 h, 37 Bq/well [3H]thymidine were added dimer exchange on DNA (16). and incorporated radioactivity measured 18 h later. For MLRs, LN T cells purified as described above (2 ϫ 105/well) were Until recently, seven NF- B inhibitor proteins were known: ϫ 5 ␣  ␥ cultured with irradiated spleen cells (2 10 /well) from syngeneic WT I B , I B ,I B ,I B , Bcl-3, p100, and p105 (reviewed in Ref. mice or allogeneic BALB/c mice for 48 h and [3H]thymidine incorporation 17). This family expanded with the identification of IBNS (7) and determined as above. IB (18–20). IBNS and IB show more sequence homology to Measurement of cytokine production each other and to Bcl-3 than to IB␣ and, like Bcl-3, these proteins appear to be nuclear rather than cytoplasmic. IB (21), Bcl-3 (22, Cytokine production was induced as above except 5 ϫ 105 LN T cells were 23), and IBNS (24, 25) have been shown to regulate promoters of used/well. After 24 h, supernatants were collected for multiplex cytokine cytokines controlled by NF-B and to be involved in inflammatory analysis (Luminex) as described (29) but using an anti-Mouse MultiCyto- kine Reporter. Concentrations were calculated using Bio-Plex Manager 3.0 responses. Their mechanisms of action are different from that of software with a five-parameter curve-fitting algorithm applied for standard IB␣ because their expression is transcriptionally regulated and curve calculations. more restricted in cell expression patterns (7, 18–21, 26, 27) unlike NK cell function and cytokine production the more ubiquitous NF-B and other IB family members. These IBNS, Bcl-3, and IB properties can narrow the action of Splenic NK cells (purity Ͼ75% as determined by anti-mouse CD49b/ NF-B providing an additional layer of specificity control. For Pan-NK DX5 (BD Biosciences) staining) were enriched using a SpinSep  Mouse NK Cell Enrichment kit (StemCell Technologies). NK cytotoxicity example, ligands such as TNF or IL-1 bind distinct receptors on was analyzed by a 51Cr-release assay (30) with mouse YAC-1 targets. a cell but use the same NF-B signaling pathways to induce dif- Enriched NK cells at 5 ϫ 105 cells/well in complete DMEM medium ferent sets of genes. Such differential gene expression can be ex- supplemented with 500 U/ml rat IL-2 were stimulated with 1 g/ml phy- plained by the fact that, in addition to activating NF-B, receptor tohemagglutinin for 3 days. Supernatants were analyzed for IFN-␥ using Downloaded from ␥ triggering may induce a particular IB family member, which dif- the Mouse IFN- OptEIA set (BD Biosciences). The sensitivity of the assay was 31.3–2000 pg/ml and results are shown as the mean from du- ferentially controls NF- B binding at specific gene promoters. We plicate wells. show in this article for the first time, using targeted gene disrup- tion, that IBNS has such a function in regulating cytokine pro- Adoptive transfer of bone marrow (BM) cells duction during T cell activation. BM cells from 2- to 3 mo-old WT or IBNS KO mice were depleted of T and B cells by CD4Ϫ, CD8Ϫ, and B220-negative separation using magnetic http://www.jimmunol.org/ Materials and Methods beads. A total of 2 ϫ 106 cells were injected i.v. into irradiated (700 rad) B6 Ly5.1 mice. Recipient mice were sacrificed at 6 wk and thymus, LN, ϩ Generation of I BNS knockout (KO) mice spleen, and peripheral blood cells analyzed for donor-derived Ly5.2 cells. Ϫ Ϫ Donor-derived Ly5.1 IAb LN T cells were prepared by magnetic bead An I BNS BAC clone was obtained from Incyte Genomics and fragments ϩ ϩ ϩ generated by PCR from the BAC clone used to generate a KO construct in negative selection and then sorted for Ly5.2 , CD4 , and CD8 T cells the pPNTloxPneo vector. 129/sv W4 embryonic stem (ES) cells were trans- and proliferation assays performed as described above. fected and rES cell lines identified by Southern analysis. Potential founders Transfection analysis were crossed to C57BL/6 mice and positive offspring bred to generate homozygous KO animals. For cognate Ag stimulation, KO mice were Transfections were performed using an Amaxa Nucleofector II and the ϩ/ϩ Ϫ/Ϫ b crossed to N15TCRtg RAG-2 H-2 mice (28). F1 and F2 offspring Nucleofector kit with modifications for use with EL4 cells. A total of 100 by guest on October 2, 2021 ϩ/ϩ Ϫ/Ϫ b of the IBNS KO ϫ N15TCRtg RAG-2 H-2 mice were inter- l of solution T are added to 1.4 ϫ 106 cells and plasmids (1 g each ϩ/Ϫ Ϫ/Ϫ ϩ/ϩ ϩ/Ϫ crossed and N15TCRtg RAG-2 IBNS ,IBNS ,or expression plasmid, 0.1 g of pRLnull transfection efficiency control and Ϫ/Ϫ ϩ/Ϫ ϩ/Ϫ ϩ/ϩ ϩ/Ϫ IBNS or N15TCRtg RAG-2 IBNS ,IBNS ,or pcDNA3.1 to total 5 g of DNA) added subsequently. After transfection Ϫ/Ϫ IBNS animals used as indicated. No difference has been observed (program O-17), cells were cultured 16–18 h before determination of ϩ/ϩ ϩ/Ϫ between IBNS or IBNS mice. Indicated biochemical experiments luciferase activity using the Dual-Luciferase Reporter Assay System ϩ/ϩ Ϫ/Ϫ b were performed using extracts from N15TCRtg RAG-2 H-2 ani- (Promega). mals. Animals were used under protocols reviewed and approved by the Dana-Farber Cancer Institute Animal Care and Use Committee. Plasmids Abs and flow cytometric analysis Plasmids for transfections: pmoIL-2–2kLuc, empty vector control and pUC00Luc (31) (all provided by S. Miyatake, Tokyo Metropolitan Institute mAbs used: R-PE- or PE-Cy5-anti-mouse CD4 (H129.19), FITC- or PE- of Medical Science, Tokyo); transfection efficiency control plasmid, pRL Cy5-anti-mouse CD8␣ (53-6.7), PE-Cy5-anti-TCR C (H57-597), PE-anti- null (Promega), pCDM8-p50, pcDNA3.1-c-Rel, pcDNA3.1-IBNS, CD25 (PC61), FITC- or PE-anti-V 5.1, 5.2 (MR9.4), PE-anti-B220 pcDNA3.1 (Invitrogen Life Technologies), pmoIL-2–2kLuc-mNF-B, and (RA3-6B2), PE-anti-NK1.1 (PK136), biotin-anti-pan NK (DX5) or PE- pmoIL-2–2kLuc-mCD28RE. The mutations of the NF-B and CD28RE Cy5-streptavidin (BD Pharmingen). For T cell V repertoire analysis, a V sites were performed with the QuickChange II Site-Directed Mutagenesis TCR screening panel (BD Pharmingen) was used. In vitro stimulations kit (Stratagene). The antisense primers for mutation in the NF-B and were performed with plate-bound purified anti-CD3 mAb 145-2C11 and CD28RE sites were 5Ј-GGGCTAACCCGACCAAGAGGCTTTTCACCT anti-CD28 clone 37.51 (eBioscience). AAATCCATTCAG-3Ј and 5Ј-CAGTGTATGGGGGTTTAAAGCCATT CCAGAGAGTCATCAGA-3Ј, respectively. The boldface letters in the Proliferation assays primers indicate the residues mutated. Mutations were confirmed by DNA sequence analysis. Total thymocytes, lymph node (LN) T cells prepared as I-Ab-negative cells ϩ ϩ by magnetic bead separation, or FACS-sorted CD4 or CD8 LN T cells DNA-binding assays from IBNS KO or WT littermate mice were used for proliferation assays. Thymocytes (1 ϫ 106/well) or LN T cells (2 ϫ 105/well) were cultured for Nuclear extracts (32) were prepared from N15TCRtgϩ/ϩRAG-2Ϫ/ϪH-2b 48 h in complete RPMI 1640 medium alone, plate-bound anti-CD3 (5 thymocytes or LN T cells (see Fig. 9, C and E) or from N15TCRtgϩ/Ϫ g/ml), anti-CD3 (5 g/ml) plus anti-CD28 (10 g/ml), Con A (2 g/ml) RAG-2Ϫ/Ϫ IBNSϩ/ϩ, ϩ/Ϫ,orϪ/Ϫ mice (Fig. 9D). Mice (4–11 wk) were plus irradiated wild-type (WT) spleen cells (2 ϫ 105/well), or PMA (50 untreated or injected i.v. with 24 g of VSV8. DNA probes containing the ng/ml) plus ionomycin (200 ng/ml). Where indicated recombinant human IL-2 promoter NF-B and CD28RE region were prepared by PCR using (rh) IL-2 (50 U/ml) was added. [3H]Thymidine (37 Bq/well) (PerkinElmer) 5‘Bio-AAACTGCCACCTAAGTGTGG3Ј and 5Ј-Bio-TTCCTCTTCTGA was added for the last 18 h of culture, and incorporated radioactivity TGACTCTC-3Ј biotinylated primers (IDT) with pmoIL-2–2kLuc, pmoIL- determined. 2–2kLuc mNF-B and pmoIL-2–2kLuc mCD28RE as templates, For peptide-induced proliferation of N15TCRtgϩ/Ϫ H-2b IBNSϩ/Ϫ or respectively. Ϫ/Ϫ T cells, total thymocytes (2 ϫ 105/well) or I-Ab-negative LN T cells 200 g of nuclear extracts were precleared with ImmunoPure Strepta- (2 ϫ 105/well) were incubated with irradiated (3 Krad) N15TCRtgϩ/Ϫ vidin-Agarose beads (Pierce) in HKMG buffer (10). A total of 1 gof H-2b IBNSϩ/ϩ or IBNSϩ/Ϫ spleen cells (1 ϫ 105/well) preloaded with biotinylated dsDNA probe and 10 g of poly(dI-dC) were added and The Journal of Immunology 1683
Table I. Summary of methods used to analyze gene expression data
Methods Detail Signature
1 Normalization: invariantset, Fold change ϭ 1.5 Modeling: pmonly, p value ϭ 0.07 Summary method: Liwong p call Ͼ20% 2 Rosetta resolver Fold change ϭ 2 p value ϭ 0.001 p call Ͼ20% 3 Normalization: Qspline, Fold change ϭ 1.5 Modeling: pmonly, p value ϭ 0.07 Background: RMA p call Ͼ20% Summary method: Liwong 4 Normalization: Quintile.robust, Fold change ϭ 1.2 Modeling: pmonly, p value ϭ 0.07 Background: RMA p call Ͼ20% Summary method: Medianpolish
DNA-bound proteins collected with streptavidin-agarose beads, washed three times with HKMG buffer, separated on SDS-PAGE, and identified by Downloaded from Western blotting using the ECL Western Blotting Detection System (Am- ersham Biosciences). Real-time PCR
RNA isolated by TRIzol (Invitrogen Life Technologies) from purified WT FIGURE 2. IBNS KO thymocytes and T cells display reduced prolif- or I BNS KO LN T cells was used for real-time PCR with the Applied erative capacity. A, Thymocytes (1 ϫ 106/well) were isolated from IBNS Biosystems Taq Man Universal PCR Master Mix. Taq Man Gene Expres- KO and WT animals and exposed to Con A (2 g/ml) plus irradiated WT http://www.jimmunol.org/ sion Premade probes were used for IL-2, -actin, Fbxo17, Bcl2l1, and spleen cells (2 ϫ 105/well), or PMA (50 ng/ml) plus ionomycin (200 ng/ Sh2d1a. Probes were custom ordered for Cpd and sequences are available 3 upon request. ml) for 48 h. [ H]Thymidine incorporation was measured for the last 18 h. B, Thymocyte cultures were set up as in A except that thymocytes were Microarray analysis cultured in the presence of plate-bound anti-CD3 (5 g/ml) or anti-CD3 3 b (5 g/ml) plus anti-CD28 (10 g/ml) for 48 h with [ H]thymidine incor- I-A -negative LN T cells prepared using magnetic beads were incubated in ϩ ϩ anti-CD3-(5g/ml) plus anti-CD28- (10 g/ml) coated plates for 0 or poration measured as in A. C, FACS-sorted CD4 or CD8 LN T cells ϫ 5 6 h. Four independently prepared sets of total RNA (TRIzol) were hybrid- from I BNS KO and WT animals were plated at 2 10 /well in the ized to standard Affymetrix GeneChip MouseGenome 430A 2.0 arrays in presence of plate-bound anti-CD3 (5 g/ml) or anti-CD3 (5 g/ml) plus the Beth Israel Deaconess Medical Center Genomics Center (Beth Israel anti-CD28 (10 g/ml) for 48 h with [3H]thymidine incorporation measured by guest on October 2, 2021 .p Ͻ 0.01 ,ءء p Ͻ 0.05 and ,ء .Hospital, Boston, MA). The quality of each array was tested using the Affy as in A
FIGURE 1. Targeted disruption of IBNS does not disturb numbers and subsets of thymocytes, LN, or splenic T and B cells. A, Schematic of wild-type IBNS gene, targeting construct and mutated allele. The probe used for Southern analysis of mutated ES cells and genotyping of mice is indicated. This probe gives a WT band of 8.75 kb and a mutated band of 5.75 kb upon DNA digestion with EcoRV. B, Genomic Southern analysis of EcoRV-digested tail DNA from WT (ϩ/ϩ), IBNS heterozygous (ϩ/Ϫ), and IBNS KO (Ϫ/Ϫ) mice. C, Analysis of cell numbers for total thymocytes and DN, DP, CD4ϩ, and CD8ϩ subpopulations in IBNS KO and WT animals. Numbers were determined on thymic cell suspensions (total number) or by FACS analysis after staining with anti-CD4 and anti-CD8 mAbs to determine the percentage of cells in the DN, DP, CD4ϩ, and CD8ϩ subpopulations. Numbers for LN and spleen were determined in the same way with B220ϩ cells classified as B cells. Animals were 6–10 wk of age. 1684 IBNS REGULATION OF T CELL CYTOKINES
FIGURE 3. IBNS KO T cells exhibit reduced responses to cognate Ag and to allogeneic MHC. A, I-Ab-negative LN T cells (2 ϫ 105/well) from N15TCRtg H-2b IBNS KO and N15TCRtg H-2b WT littermate LN T cells were assayed for proliferation to irradiated N15TCRtg H-2b WT littermate splenocytes (1 ϫ 105/well) preloaded with the indicated concentrations of VSV8. Proliferation for 48 h was determined as described in the legend for Fig. 2A. B, MLRs were performed with I-Ab-negative LN T cells from IBNS KO and WT animals and irradiated spleen cells from IBNS WT littermates or .p Ͻ 0.05 ,ء .BALB/c mice. Cultures were continued for 72 h and proliferation determined as described in the legend for Fig. 2A
package developed by Gauiter (33) considering variation in the percentage born in the expected Mendelian ratios and homozygous KOs ap- of presence calls, background, RNA degradation and scaling factors. High- pear healthy for more than 2 years. Downloaded from quality arrays were normalized and analyzed using the four different meth- As we had cloned IBNS from the thymus (7), we began anal- ods listed in Table I. All signature genes were obtained on the basis of presence calls, fold change (FC) and unpaired Student t test ( p-value) as ysis on the thymus and T cells. Surprisingly, there are no signif- shown in Table I. To reduce false-positive results, the final signature was icant differences in the thymus, LN, or spleen in terms of cell obtained by taking into consideration the genes that are identified as numbers or distribution of major subpopulations (Fig. 1C). In ad- differentially expressed by at least three methods. The unsupervised dition, the distribution of Vs expressed on peripheral CD4ϩ and ϩ two-dimensional clustering was performed on these signature genes us- http://www.jimmunol.org/ ing the DCHIP clustering utility (34). CD8 T cells is similar between WT littermates and KO animals as analyzed with 15 different anti-V Abs (data not shown). Thus, Results no overt T cell developmental or selection phenotype was IBNS KO mice observed. To examine the role of IBNS in thymocyte development, we created mice with a targeted disruption in the IBNS gene. The I BNS KO T cells are impaired in proliferation 4.8-kb deleted region contains exons I through V and 127 bp of the We tested proliferation of the IBNS KO T cells and thymocytes 153 bp exon VI, removing 208 of the 327 aa of IBNS as well as as induced with plate-bound anti-CD3, anti-CD3 plus anti- 2.8 kb upstream of exon I (Fig. 1A). Southern blot genotyping of CD28, Con A, or PMA plus ionomycin. IBNS KO thymocytes by guest on October 2, 2021 tail DNA is shown in Fig. 1B.IBNS KO mice appear grossly proliferate less in response to Con A than WT (Fig. 2A). However, normal and breed well; WT, heterozygous, and KO offspring are PMA plus ionomycin induces nearly equivalent proliferation in
FIGURE 4. Microarray and real-time PCR analysis of IBNS KO T cells. A and B, Heat maps of expression data of clustered samples in columns and clustered genes in rows for IBNS KO and WT LN T cells. The pseudocolor representation of gene expression ratios is shown in the scale below. The hierarchical clustering was performed for genes having a fold change of Ͼ1.2, p value 0.007 in at least three analytical methods. A, The hierarchical clustering of eight genes differentially expressed comparing nonstimulated WT and IBNS KO T cells. B, The hierarchical clustering of 31 genes differentially expressed comparing WT and IBNS KO T cells stimulated for 6 h with plate-bound anti-CD3 ϩ anti-CD28 mAbs. C, Real-time PCR analysis on RNA from WT and IBNS KO T cells of five genes determined to be differentially expressed in B. The Journal of Immunology 1685
Table II. Genes expressed differentially between WT and IBNS KO LN T cells
Gene Symbol Description Accession Fold Change p Value
Genes expressed differentially between WT and IBNS KO LN T cells at 0 h AY078069 NF-B inhibitor (IBNS) NM_172142 Ϫ5.1 0.00199625 Sbsn Suprabasin NM_172205 Ϫ4.43 0.006232 2610042L04Rik RIKEN cDNA 2610042L04 gene NM_001024713 Ϫ2.91 0.00545575 Hcst Hematopoietic cell signal transducer AF172930 Ϫ2.43 0.00421775 Ctla4 CTL-associated protein 4 NM_009843 Ϫ2.06 0.00999116 Tyrobp TYRO protein tyrosine kinase-binding protein NM_011662 Ϫ2.04 0.00743101 AW551225 Expressed sequence AW551225 NM_198109 Ϫ1.99 0.004768 Fbxo17 F-box only protein 17 NM_015796 5.881 0.001388 Genes expressed differentially between WT and IBNS KO LN T cells at 6 h postactivation with anti-CD3⑀ ϩ anti-CD28 Ppic Peptidylprolyl isomerase C NM_008908 Ϫ2.21 0.012656 Pep4 Peptidase D NM_008820 Ϫ2.57 0.001669 Irf8 IFN regulatory factor 8 BC005450 Ϫ2.16 0.003051 Ifitm2 IFN-induced transmembrane protein 2 NM_030694 Ϫ2.15 0.036444 Tspan13 Tetraspanin 13 NM_025359 Ϫ2.06 0.018842 Tnfsf11 TNF (ligand) superfamily, member 11 NM_011613 Ϫ2.18 0.028869 Hcst Hemopoietic cell signal transducer AF172930 Ϫ2.36 0.02854 Il1r2 IL-1 receptor, type II NM_010555 2.18 0.017086 Ksr Kinase suppressor of ras 1 AW492498 Ϫ2.09 0.005676 Downloaded from Gp49a Glycoprotein 49 A U05264 2.11 0.000293 Bcl2l1 Bcl2-like 1 BM228788 Ϫ2.06 0.014994 Fbxo17 F-box protein 17 NM_015796 2.3 0.00359 Rora RAR-related orphan receptor ␣ BC003757 Ϫ2.04 0.024768 Csrp1 Cysteine and glycine-rich protein 1 BF124540 2.27 0.064419 Plagl1 Pleiomorphic adenoma gene-like 1 AF147785 Ϫ1.9 0.000192 Csf2 CSF2 (granulocyte-macrophage) X03019 2.02 0.027682 http://www.jimmunol.org/ Igl-V1 Ig chain, variable 1 AK008551 Ϫ3.55 0.044471 AY078069 NF-B inhibitor IBNS AW495632 Ϫ10.91 0.011061 AA855988 Transcribed locus AA855988 Ϫ1.69 0.001034 Map4k1 MAPK kinase kinase kinase 1 BB546619 Ϫ2.19 0.000064 5730469M10Rik RIKEN cDNA 5730469M10 gene AV332575 Ϫ1.87 0.003323 Sh2d1a SH2 domain protein 1A NM_011364 Ϫ3.49 0.002479 Il2 IL-2 AF065914 Ϫ2.59 0.062586 Zfpm1 Zinc finger protein, multitype 1 AA014267 Ϫ1.91 0.002094 LOC544986 Hypothetical protein LOC544988 BM195235 Ϫ2.4 0.001473 Cpd Carboxypeptidase D AW550842 Ϫ5.5 0.040013
Igsf3 Ig superfamily, member 3 BG075582 2.04 0.000554 by guest on October 2, 2021 AI850995 Glucose-fructose oxidoreductase domain containing 1 AV220135 Ϫ1.73 0.00553 Hnrpll Heterogeneous nuclear ribonucleoprotein L-like AI503931 Ϫ1.76 0.003628 9630010G10 RIKEN cDNA 9630010G10 gene BI076710 Ϫ2.08 0.027778
WT and IBNS KO thymocytes. IBNS KO thymocytes (Fig. 2B) T cells require VSV8 at ϳ2 logs higher concentration for prolif- as well as CD4ϩ and CD8ϩ LN T cells (Fig. 2C) exhibit reduced eration equivalent to WT T cells. proliferation to anti-CD3 and anti-CD3 plus anti-CD28. The IBNS KO T cells were tested for recognition of allogeneic proliferation in KO thymocytes is more severely reduced than in MHC in MLR. The LN T cells of the IBNS KO animals do not the mature KO T cells. We previously demonstrated that IBNS is respond to syngeneic splenocytes, but are activated by splenocytes induced by TCR cross-linking and not by agents that do not act from MHC H-2d mice, although to a significantly reduced extent directly through TCR ligation (7). The fact that PMA plus iono- compared with LN T cells from WT littermates (Fig. 3B). mycin bypasses the IBNS deficiency is consistent with this com- bination acting downstream of the TCR beyond the point at which Microarray analysis of I BNS KO T cells IBNS is required. Con A, by cross-linking CD3 glycans (35), The genetic consequences of IBNS deletion on T cell activation likely acts via the TCR complex and, perhaps, additional surface were determined using four independent microarray analyses. Un- receptors and therefore, is impaired like the anti-CD3 Ϯ anti- stimulated KO and WT T cells segregate completely and are dis- CD28 Ab cross-linking stimulation. criminated by eight genes including IBNS (AY078069)(Fig. 4A and Table II “Genes expressed differentially between WT and I BNS KO T cells exhibit reduced responses to cognate and IBNS KO LN T cells at 0 h”). This low number is not surprising allogeneic Ags because we observe normal development of T cells in the IBNS To examine IBNS KO T cell responses to cognate Ag, IBNS KO mouse and because IBNS expression is not constitutive but KO mice were crossed with N15TCRtgϩ/ϩRAG-2Ϫ/ϪH-2b mice rather induced upon TCR triggering (7). In addition, the expression (hereafter termed N15TCRtgϩ/ϩ mice) (28). N15TCRtgϩ/ϩ mice levels of Hcst, Fbxo17, and Sbsn are more similar in the IBNS express a TCR recognizing a vesicular stomatitis nucleoprotein KO and WT T cells after activation. Expression of Tyrobp in the octapeptide (VSV8) in the context of H-2Kb (36). Thymocytes and IBNS KO is decreased ϳ2-fold at both zero and 6 h while the LN T cells purified from N15TCRtgϩ/ϪRAG-2ϩ/ϪIBNSϪ/Ϫ and level of IBNS expression is 5-fold down at 0 h and 10.9-fold N15TCRtgϩ/ϪRAG-2ϩ/ϪIBNSϩ/ϩ littermate animals were down at 6 h. Expression of IBNS in the unstimulated WT LN T tested for proliferation in response to peptide-loaded, irradiated cells may result from manipulations conducted during cell IBNSϩ/ϩ spleen cells (Fig. 3A). The N15TCRtgϩ/Ϫ IBNS KO separation. 1686 IBNS REGULATION OF T CELL CYTOKINES Downloaded from
FIGURE 5. IBNS KO NK cells exhibit efficient cell killing function but are deficient in IFN-␥ production. A, Enriched NK cells from IBNS KO and WT spleens were assayed for cytotoxicity on YAC-1 target cells using a 51Cr-release assay. B, Enriched NK cells from IBNS KO and WT spleens were cultured in the presence of 500 U/ml rat IL-2 and 1 g/ml PHA for 3 days. IFN-␥ production was determined by ELISA on the su- http://www.jimmunol.org/ pernatants of these cultures.
FIGURE 6. IBNS KO thymocytes and LN T cells are defective in IL-2 ␥ Analyses of the activated IBNS KO and WT T cells yielded a and IFN- cytokine production but can respond to exogenous IL-2. A, ␥ final signature containing 30 genes (Fig. 4B and Table II “Genes Production of IL-2 and IFN- by I BNS KO and WT thymocytes stimu- lated 24 h as indicated. Culture conditions are as described in Materials expressed differentially between WT and IBNS KO LN T cells at ϩ and Methods. Cytokine production was assayed by Luminex. B, IL-2 and 6 h postactivation with anti-CD3 anti-CD28”). Only I BNS, IFN-␥ cytokine production by FACS-purified CD8ϩ LN T cells from Hcst, and Fbxo17 were present in the final signature of both resting IBNS KO and WT animals stimulated as indicated was determined as in by guest on October 2, 2021 and stimulated T cells. The 30 genes in the six hour signature are A. C, I-Ab-negative LN T cells from IBNS KO and WT animals were distributed over 15 different chromosomes with six located on stimulated as indicted with (ϩIL-2) and without (ϪIL-2) addition of 50 chromosome 7; IBNS maps to 7A3 as do Hcst, Fbxo17, and U/ml rhIL-2. Proliferation was determined as described in the legend for .p Ͻ 0.01 ,ءء p Ͻ 0.05 and ,ء .Map4k1 (chromosome locations from National Center for Biotech- Fig. 2A nology Information). Pep4 at 7B1 and Ifitm2 at 7F5 are more dis- tant. The significance of the dysregulation of these genes and their location on the same chromosome as IBNS is undetermined. IBNS may play a role in regulating these genes but we cannot I BNS KO thymocytes and T cells are impaired in cytokine exclude the possibility that the IBNS gene disruption alters tran- production scription in some global manner around that position on Because primary T cell proliferation depends on cytokine produc- chromosome 7. tion, particularly IL-2, we used cytokine multiplex analysis to Tyrobp and Hcst encode adapter proteins that participate in measure cytokine production by IBNS KO and WT littermate T forming activating receptors on NK cells (reviewed in Ref. 37). cells and thymocytes in response to plate-bound anti-CD3, anti- Given their altered expression, we tested the cytotoxic capacity of CD3 plus anti-CD28, Con A and PMA plus ionomycin. Upon IBNS KO NK cells. IBNS KO NK cells killed Yac-1 targets as TCR cross-linking, IL-2 and IFN-␥ production are reduced in efficiently as NK cells from IBNS WT littermates (Fig. 5A). Fur- IBNS KO animals (Fig. 6, A and B). IL-2 production is reduced thermore, flow cytometric analysis demonstrated equivalent levels in total thymocytes (Fig. 6A), CD4ϩ (data not shown) and CD8ϩ of NKG2D on the surface of IBNS KO and WT NK cells (data LN-derived IBNS KO T cells (Fig. 6B). IFN-␥ production is not shown). Thus, a 2-fold reduction in Hcst mRNA does not affect reduced in IBNS KO CD8ϩ T cells and thymocytes undergoing expression of NKG2D or the killing function of NK cells. In ad- similar stimulation. Under these conditions, IBNS KO and WT dition, numbers of NK cells are equivalent between the WT and CD4ϩ T cells do not produce detectable IFN-␥. With PMA plus KO mice. IBNS KO NK cells are defective in IFN-␥ production, ionomycin, IL-2 production was almost equal between IBNS KO however, when cultured in vitro (infra vide and Fig. 5B). and WT thymocytes and T cells (data not shown) indicating that To confirm these differences in gene expression, we performed KO cells can produce more IL-2 with stimuli bypassing the TCR. real-time PCR analysis of Fbxo17, Sh2d1a, Bcl2l1, Il2, and Cpd. This observation parallels the effect of PMA plus ionomycin on The real-time PCR results substantiate higher expression of proliferation as described above. IL-2 production by N15TCRtgϩ/Ϫ Fbxo17 and lower Sh2d1a, Bcl2l1, Il2, and Cpd in IBNS KO T RAG-2ϩ/ϪIBNSϪ/Ϫ and N15TCRtgϩ/ϪRAG-2ϩ/ϪIBNSϩ/ϩ cells (Fig. 4C). Furthermore, as discussed below, the IL-2 defi- LN T cells stimulated with VSV8-loaded APCs shows that cognate ciency is confirmed at the protein level in our cytokine assays (Fig. peptide stimulation of IBNS KO T cells results in much lower 6, A and B). IL-2 production compared with WT T cells (Fig. 7). The Journal of Immunology 1687
notype observed in IBNS KO animals is cell autonomous and not dependent on the environment in which the T cells mature.
IBNS transfection increases IL-2 promoter transcription The IL-2 promoter is controlled by a number of transcription fac- tors including NF-B family members (38–41). The major NF-B contributor to induction of IL-2 is c-Rel (42). Indeed, the c-Rel and IBNS KO mice have similar phenotypes in that the immune sys- tems appear normal but the T cells make reduced IL-2 upon acti- vation (8, 9). Because the IBNS protein has no DNA-binding FIGURE 7. IBNS KO LN T cells produce less IL-2 than WT cells upon cognate Ag activation. N15TCRtgϩ/Ϫ RAG-2Ϫ/Ϫ WT or IBNSϪ/Ϫ domain, its affect must be mediated through interaction with a LN T cells were expose to the indicated concentrations of VSV8 on APCs DNA-binding protein and the similar phenotypes of the c-Rel and and IL-2 in the culture supernatant assayed by ELISA. IBNS KO mice suggest that c-Rel and IBNS might interact to regulate IL-2. To examine the role of IBNS and c-Rel in control of the IBNS KO T cells are capable of responding to IL-2. Exoge- IL-2 promoter, we used transfection assays with a plasmid in nous IL-2 reduces the disparity in proliferation between IBNS which the IL-2 promoter drives a luciferase reporter (Fig. 9A). KO and WT LN T cells (Fig. 6C). Proliferation of both IBNS KO EL4 cells were transfected with the IL-2 promoter vector only and WT T cells is increased by IL-2 addition, but that of the or in combination with IBNS- or c-Rel-expressing constructs. Downloaded from IBNS KO cells is more strongly enhanced. By flow cytometry, no Because we previously determined that IBNS binds p50 (7), obvious difference in expression of CD25 was observed in IBNS we also tested the effect of p50 transfection either alone or in KO mice (data not shown). combination with IBNS and c-Rel. IL-2 promoter transcription To test whether the observed defects are intrinsic to IBNS KO is increased upon cotransfection with either IBNS or c-Rel T cells, we examined IL-2 production in thymocytes and T cells (Fig. 9B, upper panel). Cotransfection of IBNS plus c-Rel from BM chimeric animals created by injecting IBNS KO or WT activates transcription from the IL-2 promoter in an additive http://www.jimmunol.org/ BM into irradiated B6Ly5.1 recipients. Repopulation of thymus fashion. Cotransfection of p50 did not result in altered tran- and LN 6 wk after injection was equivalent for IBNS KO and WT scription of the IL-2 promoter either alone or with IBNS or (Fig. 8A). However, LN T cells from the IBNS KO BM chimeras c-Rel (data not shown). exhibited proliferation defects in vitro, which were largely restored A canonical and a noncanonical NF-B site, the latter termed a by addition of exogenous IL-2 (Fig. 8B). Thus, the T lineage phe- CD28 response element (CD28RE), lie within 300 bp of the IL-2 by guest on October 2, 2021
FIGURE 8. Defect in proliferation and IL-2 production is intrinsic to IBNS KO LN T cells. A, Analysis of radiation chimeras constructed with BM from IBNS KO or WT litter- mate animals. Ly5.2ϩ donor-derived cells from the thymus (upper)orLN (lower) of chimeric animals were an- alyzed for T cell reconstitution using CD4 and CD8 as markers. B, I-Ab- negative LN cells isolated from IBNS KO or WT chimeras were as- sayed for proliferation in the presence and absence of exogenous IL-2 as de- scribed in Materials and Methods. 1688 IBNS REGULATION OF T CELL CYTOKINES Downloaded from http://www.jimmunol.org/ by guest on October 2, 2021
FIGURE 9. IBNS stimulates transcription from the IL-2 promoter and is dependent on the NF-B site for activity and p50-independent asso- ciation with IL-2 promoter DNA. A, Schematic diagram of the IL-2 promoter/luciferase reporter construct used for transfection analysis with the sequence of the NF-B and CD28RE sites and mutations therein delineated. B, Upper panel, Fold increase in IL-2 promoter-controlled luciferase activity in transfections comparing the IL-2 promoter construct alone (pro-IL-2) or in combination with a c-Rel-expressing plasmid (ϩc-Rel), an IBNS-expressing plasmid (ϩIBNS) or both IBNS and c-Rel (ϩIBNS ϩc-Rel). Results are a compilation of 14, 18, and 10 assays, respectively. Throughout the analysis, all measurements were normalized by dividing by the corresponding control (pro-IL-2) measurement. Statistical analysis was performed applying the unpaired Student t test with unequal variance. Middle and bottom panels, Fold increase in IL-2 promoter-controlled luciferase activity in transfections using the CD28RE- (mCD28RE) or NF-B- (mNF-B) mutated IL-2 promoter constructs. Results are a com- pilation of seven assays each. Statistical analysis was performed applying the unpaired Student t test with unequal variance comparing the result of ,p Ͻ 0.001 or ‚, p Ͻ 0.05 comparing the pro-IL-2 ,ء .the transfections using the mutated constructs with transfections with the pro-IL-2 promoter mCD28, or mNF-B reporter construct alone with the reporter construct ϩ c-Rel, ϩ IBNS, or ϩ c-Rel ϩ IBNS. , p Ͻ 0.001 comparing ϩ c-Rel, ϩ IBNS and ϩ IBNS ϩ c-Rel transfections using the mCD28 or mNF-B promoter (middle and bottom panels, respectively) with the equivalent transfections on the pro-IL-2 promoter (upper panel). C, Western blot analysis of DNA pulldowns on 100 g of lysates from control and VSV8-activated LN T cells isolated from N15TCRtgϩ/ϩ RAG-2Ϫ/Ϫ H-2b animals with no injection (Ϫ) or 1 h after i.v. VSV8 injection (ϩ). Lysate lanes (two left lanes) show total nuclear extracts (8 g in all lysate lanes) used for DNA pulldown and WT, mNF-B and mCD28RE lanes show the presence of IBNS and p50 in DNA pulldowns using the WT, NF-B mutant, or CD28RE mutant IL-2 promoter DNA probes. D, Western blot analysis of DNA pulldowns on 210 g of lysates from LN T cells isolated from N15TCRtgϩ/Ϫ RAG-2Ϫ/Ϫ H-2b IBNSϩ/ϩ or ϩ/Ϫ (WT) or N15TCRtgϩ/Ϫ RAG-2Ϫ/Ϫ H-2b IBNSϪ/Ϫ (KO) animals with no injection (Ϫ) or 2 h after i.v. VSV8 injection (ϩ). Lysate lanes (four left lanes) show the presence of IBNS, p50, and c-Rel in the total nuclear extracts used for DNA pulldowns with the WT IL-2 promoter probe (four right lanes). E, Western blot analysis of DNA pulldowns on 180 g of lysates from LN T cells isolated from N15TCRtg RAG-2Ϫ/Ϫ mice 0, 0.5, 1, or 2 h after i.v. VSV8 injection. The lysate lanes (four left lanes) show the total nuclear lysates used for the DNA pulldown with the WT IL-2 promoter probe (four right lanes). promoter (43–45). Specificities of the IBNS and c-Rel activities on the IL-2 promoter (Fig. 9B, bottom panel); c-Rel regulation of were investigated using individual NF-B and CD28RE mutants the NF-B-mutated promoter is intact. Conversely, mutation of (Fig. 9A). Mutation of the NF-B site disrupts the effect of IBNS the CD28RE blocks activation of transcription by c-Rel while The Journal of Immunology 1689 induction of transcription by IBNS is not affected (Fig. 9B, mid- dle panel). Thus, as expected, c-Rel acts on the CD28RE element. IBNS, however, acts through the NF-B site. IBNS association with the IL-2 promoter To detect binding of transcription factors to the IL-2 promoter, a biotinylated DNA-binding assay was used. A 109-bp probe (bp Ϫ134 to Ϫ242) (numbering according to Ref. 46) was added to nuclear lysates extracted from thymus and LN cells and DNA- bound proteins detected by Western blotting after DNA precipita- tion. Samples were prepared from N15TCRtgϩ/Ϫ IBNS KO mice and N15TCRtgϩ/Ϫ WT littermates after i.v. injection with VSV8 peptide for various times. The IBNS/DNA pulldown matches the specificity demonstrated in transfection assays; one hour following VSV8 injection, nuclear IBNS can be precipitated with WT or FIGURE 10. Real-time PCR analysis of the kinetics of IL-2 transcrip- CD28RE mutant (mCD28RE) but not with NF- B mutant (mNF- tion in LN T cells from I BNS KO and WT littermates. RNA was isolated B) DNA (Fig. 9C). Note that although the level of I BNS in the from purified LN T cells activated in vitro with plate-bound anti-CD3 ϩ total nuclear lysate is below detectable levels, significant amounts anti-CD28 for the indicated times and quantified by real-time PCR. of IBNS are pulled down with IL-2 promoter DNA and this is dependent on the NF-B site. p50 shows the same specificity of Downloaded from IL-2 binding as IBNS; hence, there is no binding to the mNF-B As IBNS exhibits no DNA binding domain and as little or no DNA. However, p50 is present in the lysates in the absence of c-Rel or p50 is present on the IL-2 promoter DNA at the later times VSV8 stimulation and binds to the IL-2 promoter DNA 1 h after where IBNS is pulled down by the DNA (Fig. 9, D and E), other VSV8 stimulation as well as in lysates from unstimulated T DNA-binding proteins likely interact with IBNS to form a com- cells (Fig. 9C). Two hours postinjection, the amount of p50 plex on the IL-2 promoter DNA. pulled down by the IL-2 promoter is barely detectable even http://www.jimmunol.org/ though significant p50 is present in the lysate (Fig. 9D). Sim- Discussion ilarly, at 2 h c-Rel is present in the lysates of both WT and Our initial characterization of IBNS showed a correlation be- IBNS KO LN T cells, but is bound to the IL-2 promoter only tween thymocyte negative selection and induction of IBNS (7). in control, not activated, lysates (Fig. 9D). However, c-Rel We concluded that IBNS played a role in thymic negative selec- binds to the IL-2 promoter in lysates prepared 30 min after tion and created a targeted gene disruption to further characterize VSV8 treatment (Fig. 9E); this binding is decreased or com- this gene. We anticipated greater numbers of thymocytes and T pletely gone (Fig. 9E) after1hasdescribed further below. cells in IBNS KO animals compared with WT littermates and that Although there appears to be slightly less p50 in the IBNS KO the IBNS KO animals might suffer the effects of self-reactive T by guest on October 2, 2021 than in the WT (Fig. 9D), the DNA binding patterns appear cells because these would not be deleted if negative selection was nearly identical between the IBNS KO and WT lysates. Like- disrupted. However, this was not the case. Analysis of V expres- wise, the binding pattern exhibited by c-Rel appears very sim- sion in the IBNS KO T cell population indicated no effect of the ilar in IBNS KO vs WT lysates (Fig. 9D). As expected, the disruption of IBNS on negative selection and IBNS KO animals IBNS KO lysate has no IBNS protein (Fig. 9D). show no ongoing autoimmune processes or obvious developmental A comparison of the kinetics of p50, c-Rel, and IBNS expres- abnormalities in the immune system. Thus, if IBNS does play a sion in LN T cells and IL-2 promoter binding after TCR triggering role in the process of negative selection, its absence must be com- is provided in Fig. 9E. Lysates were prepared 0, 0.5, 1, or 2 h after pensated for by other mechanisms. in vivo VSV8 treatment of N15TCRtgϩ/ϩ mice. p50 and c-Rel in However, the proliferation of IBNS KO thymocytes and LN T the lysate increase 0.5 h after VSV8 activation of T cells while cells in response to plate-bound anti-CD3, anti-CD3 plus anti- IBNS is not detected until 2 h. The amount of p50 and c-Rel CD28 as well as Con A is significantly reduced. The response to bound to the IL-2 promoter increases in parallel with their expres- alloantigen and to cognate peptide Ag is also diminished in IBNS sion level from 0 to 0.5 h but subsequently the amount bound by KO T cells. Even at high cognate peptide concentrations, the level the IL-2 promoter DNA decreases even though the level present in of proliferation in the IBNS KO LN T cells never reaches the the lysate remains constant (Fig. 9E). IBNS, in contrast, is pulled level seen in WT littermate LN T cells (Fig. 3). down with the IL-2 promoter DNA even though below detectable As IBNS alters the activity of the transcription factor, NF-B, levels in the lysate. Furthermore, the amount of IBNS pulled we used microarray analysis to assess the effect of deletion of this down with the IL-2 promoter DNA continues to increase up to at factor on global gene transcription in T cells. Because we observed least 2 h post-TCR activation. differences in activation and proliferation in vitro, we isolated What are the effects of these different IL-2 promoter DNA/ RNA from IBNS KO and WT littermate LN T cells activated by protein combinations on IL-2 transcription? Real-time PCR plate-bound anti-CD3 plus anti-CD28. Without stimulation, eight analysis shows that IL-2 message levels are equal between WT genes are differentially expressed in the IBNS KO T cells; these and IBNS KO LN T cells at 0.5 and 1 h after anti-CD3 plus include two unknown cDNAs, an F-box protein, three proteins anti-CD28 stimulation in vitro, but by 2 and 4 h transcription is with known roles in the immune system, IBNS itself and supra- 3- to 5-fold lower in the IBNS KO T cells (Fig. 10). These data basin, a protein of unknown function expressed in keratinocytes indicate that IBNS is required for IL-2 production, although (47). All genes except Fbxo17 are decreased in the IBNS KO T the kinetics of the IBNS effect appear somewhat slower than cells. those observed in Fig. 9E. This may be due to the fact that T cell At 6 h activation, 30 genes are significantly different in expres- activation in vitro is less efficient than cognate Ag T cell acti- sion between the WT and IBNS KO T cells (Table II “Genes vation in vivo. expressed differentially between WT and IBNS KO LN T cells at 1690 IBNS REGULATION OF T CELL CYTOKINES
6 h postactivation with anti-CD3 ϩ anti-CD28”). The gene with Il2 is present among the 20 genes with Ϫ1.69 to Ϫ2.59 FCs; this the greatest fold change is IBNS itself, which is 10.91-fold de- is confirmed by cytokine and RT-PCR analyses. IBNS KO T cells creased in the IBNS KO T cells. Cpd at Ϫ5.5, Igl-V1 at Ϫ3.55 as well as NK cells show a clear deficiency in IFN-␥ production. and Sh2d1a at Ϫ3.49 exhibit the next largest differences in ex- However, decreased IFN-␥ production was not apparent in the pression. Twenty genes exhibit FC of Ϫ1.69 to Ϫ2.59 including microarray analysis, probably due to the 6-h time point chosen, five unknowns, genes induced by IFN-␥, proteases, cell surface which may be too short to detect such differences. molecules, zinc finger proteins, signaling molecules and transcrip- Thus, disruption of IBNS results in altered expression of tion factors. Genes of particular interest are discussed further cytokines, transcription factors, kinases, isomerase and pepti- below. dase enzymes, and ligands for TNF (Table II “Genes expressed In the activated microarray samples, six genes exhibit FCs of differentially between WT and IBNS KO LN T cells at 6 h 2.02 to 2.3, including an IL-1 decoy receptor (Il1r2), a LIM-do- postactivation with anti-CD3 ϩ anti-CD28”). Which of these, main containing protein involved in gene regulation, cell growth, aside from the II2 gene, are direct targets of IBNS remains to development, and differentiation (Csrp1), two Ig-like cell surface be determined but this analysis serves as a rich source of further receptors (Igsf3 and Gp49a), an F-box-containing protein involved investigation into the role of IBNS and NF-B in regulation of in ubiquitylation (Fbxo17) and GM-CSF. Igsf3 (CD101) has been T cell activation. implicated as the type 1 diabetes locus, Idd10, (48) and triggering I BNS suppresses LPS-induced IL-6 production in colonic lam- of CD101 decreases activation of T cells (49). Gp49a is an inhib- ina proprial macrophages (24). Independent analysis of an I BNS itory cell surface molecule that suppresses IFN-␥ responses of T KO (25) showed that I BNS inhibited NF- B induction of IL-6 and IL-12p40. The phenotype was similar to our IBNS KO with
and NK cells (50). The Il1r2 gene is a negative regulator of T cell Downloaded from activation and thus IFN-␥ and IL-2 production (51). IBNS KO the exception of spontaneous development of chronic colitis in thymocytes and T cells produce much less IFN-␥ upon TCR trig- their mice. Kuwata et al., also observed normal T and B cell gering by anti-CD3 or anti-CD3 plus anti-CD28 (Fig. 6). Al- development and concluded that T cells in their I BNS KO though it is possible that IBNS directly acts to control the IFN-␥ proliferated normally; however their assays were performed in promoter, an alternative explanation is that the increases in Il1r2, the presence of exogenous IL-2 thus minimizing the prolifera- Igsf3 Gp49a tion difference between the I BNS KO and WT cells (Fig. 6C). ,and , all of which are inhibitory in terms of T cell http://www.jimmunol.org/ They concluded that IBNS acts to inhibit NF-B activity on activation, are the cause of the lower IFN-␥ production in the IL-6 and IL-12p40 promoters because, in their KO mice, mac- IBNS KO. GM-CSF mRNA is also increased in the IBNS KO rophages, and dendritic cells produce more of these cytokines. T cells. As opposed to Il1r2, Igsf3, and Gp49a, which appear to Our data with the IL-2 promoter are not consistent with a com- inhibit T cell activation, GM-CSF plays a proinflammatory role in parable inhibitory mechanism of action for IBNS in T cells. In the immune system. Although an increase in Il1r2, Igsf3, and contrast, IL-2 and IFN-␥ production from T lymphocytes is Gp49a should therefore decrease T cell activation, an increase in lower in our IBNS KO animals suggesting that IBNS medi- GM-CSF should have the opposite effect. As there is no evidence ates distinct activities in a variety of cell types through regu- for excessive inflammation in these mice, perhaps the pro- and lation of different promoters. anti-inflammatory changes cancel each other out. by guest on October 2, 2021 It has been shown in intestinal macrophages that IBNS binds to Cpd encodes carboxypeptidase D, an integral membrane protein p50 homodimers and reduces expression of IL-6 (24). Because expressed in many tissues that may play a role in a secretory path- IBNS coimmunoprecipitates with p50 both in our EL-4 transfec- way contributing to peptide hormone processing (52). Although tants and from lysates isolated from ex vivo cells (data not shown), the lower expression level in I BNS KO T cells was confirmed by one possibility was that IBNS bound p50 homodimers removing RT-PCR, the relevance to T cell function remains undetermined. these from the IL-2 promoter and allowing c-Rel to enter this site. Sh2d1a encodes a protein, SLAM-associated protein (SAP; where Instead, however, c-Rel binds to the IL-2 promoter DNA within SLAM is signaling lymphocyte activation molecule or CD150), 0.5 h after TCR triggering and then decreases (Fig. 9E and data not involved in regulation of Th2-type cytokines and NK cytotoxicity shown). Similarly, p50 binds the IL-2 promoter at earlier time Ϫ/Ϫ ϩ (reviewed in Ref. 53). SAP CD4 T cells produce less IL-4 points but is reduced or has disappeared by 2 h while IBNS ␥ Ϫ/Ϫ and IL-13 but more IFN- and SAP NK cells exhibit reduced binding increases from 1 to 2 h. Thus binding of IBNS to the IL-2 cytotoxicity. Such deficits are not recapitulated in our I BNS KO promoter appears to be independent of p50 and c-Rel at later time mice as they have normal NK cell cytotoxicity, express reduced points. Given that c-Rel has been shown to play a role in chromatin levels of IFN-␥ and IBNS KO CD4ϩ T cells produce as much ϩ remodeling and opening of the IL-2 promoter for transcription IL-4 and IL-13 as WT control CD4 T cells (data not shown). (57), one possibility is that p50 and c-Rel open the IL-2 promoter Thus, the reduced SAP mRNA levels do not account for the for transcription and, subsequently, IBNS along with other pro- IBNS KO phenotype. teins then maintains transcription from the promoter thus extend- Expression levels of two IFN-␥ responsive genes, Irf8 and ing IL-2 production in activated T cells. If so, transcription from Ifitm2, are reduced in IBNS KO T cells. Irf8 encodes a transcrip- the IL-2 promoter should initially be identical in WT and IBNS tion factor implicated in the immune response (reviewed in Ref. KO T cells upon TCR activation but should diverge at the point 54) and Ifitm2 is a transmembrane protein with a role identified in where IBNS plays a role in this process. Such kinetics are ob- germ cell specification (55). Other genes of note in this category served in comparing real-time PCR analysis of IL-2 transcription include the enzyme Ppic encoding cyclophilin C, the anti- between WT and IBNS KO T cells (Fig. 10). apoptotic gene Bcl2l1, Tnfsf11 encoding RANKL or TRANCE, Because IBNS has no DNA binding domain, it must interact transcription factors including zinc finger proteins (Plagl1, Zfpm1) with other protein components to associate with IL-2 promoter and the transcription factor Rora known to be involved in lym- DNA. Proteins known to affect IL-2 expression are histone phocyte development (56). All of these genes with the exception of deacetylase (Ref. 58 and reviewed in Ref. 59) and HMG-Y (60, Plagl1 have known roles in the immune system. In addition, the 61). In addition, Creb-binding protein and p300 affect NF-B- signal molecule-encoding Map4k1 and Ksr genes are decreased in driven transcription (reviewed in Ref. 59). The finding that the IBNS KO. IBNS interacts with proteins bound at the IL-2 promoter adds The Journal of Immunology 1691
further complexity to the analysis of the proteins altering tran- 23. Wessells, J., M. Baer, H. A. Young, E. Claudio, K. Brown, U. Siebenlist, and scription of this important cytokine. Identification of IBNS- P. F. Johnson. 2004. Bcl-3 and NF- B p50 attenuate lipopolysaccharide-induced inflammatory responses in macrophages. J. Biol. Chem. 279: 49995–50003. binding partners present at the IL-2 promoter is a subject of 24. Hirotani, T., P. Y. Lee, H. Kuwata, M. Yamamoto, M. Matsumoto, L. Kawase, future investigation. S. Akira, and K. Takeda. 2005. The nuclear IB protein IBNS selectively in- hibits lipopolysaccharide-induced IL-6 production in macrophages of the colonic lamina propria. J. Immunol. 174: 3650–3657. Acknowledgments 25. Kuwata, H., M. Matsumoto, K. Atarashi, H. Morishita, T. Hirotani, R. Koga, and We thank Drs. R. S. Blumberg, S. Koyasu, H.-C. Liou, and I. Schmitz for K. Takeda. 2006. I BNS inhibits induction of a subset of Toll-like receptor- dependent genes and limits inflammation. Immunity 24: 41–51. critical reading of the manuscript, Drs. L. Glimcher and E. Hwang for 26. Ohno, H., G. Takimoto, and T. W. McKeithan. 1990. The candidate proto-on- advice with DNA-binding assays, Dr. H. Kim for statistical analysis, and cogene bcl-3 is related to genes implicated in cell lineage determination and cell K. Sigrist for animal husbandry creating the knockout mouse. Dr. H.-C. cycle control. Cell 60: 991–997. Liou generously provided NF-B expression plasmids. 27. Bhatia, K., K. Huppi, T. McKeithan, D. Siwarski, J. F. Mushinski, and I. Magrath. 1991. Mouse bcl-3: cDNA structure, mapping and stage-dependent expression in B lymphocytes. Oncogene 6: 1569–1573. Disclosures 28. Ghendler, Y., R. E. Hussey, T. Witte, E. Mizoguchi, L. K. Clayton, A. K. Bhan, The authors have no financial conflict of interest. H.-C. Chang, and E. R. Reinherz. 1997. Double-positive T cell receptor-high thymocytes are resistant to peptide/major histocompatibility complex ligand-in- duced negative selection. Eur. J. Immunol. 27: 2279–2289. References 29. Chang, D. H., K. Osman, J. Connolly, A. Kukreja, J. Krasovsky, M. Pack, A. Hutchinson, M. Geller, N. Nishi, Y. Ando, et al. 2005. Sustained expansion of 1. Starr, T. K., S. C. Jameson, and K. A. Hogquist. 2003. Positive and negative NKT cells and antigen-specific T cells after injection of ␣-galactosyl-ceramide selection of T cells. Annu. Rev. Immunol. 21: 139–176. loaded mature dendritic cells in cancer patients. J. Exp. Med. 201: 1503–1517. 2. Palmer, E. 2003. Negative selection– clearing out the bad apples from the T-cell repertoire. Nat. Rev. Immunol. 3: 383–391. 30. Reche, P. A., D. B. Keskin, R. E. Hussey, P. Ancuta, D. Gabuzda, and E. L. Reinherz. 2006. Elicitation from virus-naive individuals of cytotoxic T Downloaded from 3. Weiss, A., and D. R. Littman. 1994. Signal transduction by lymphocyte antigen lynmphocytes directed against conserved HIV-1 epitopes. Med. Immun. 5: 1–14. receptors. Cell 76: 263–274. 4. Hettmann, T., J. DiDonato, M. Karin, and J. M. Leiden. 1999. An essential role 31. Tsuruta, L., H. J. Lee, E. S. Masuda, N. Koyano-Nakagawa, N. Arai, K. Arai, and for nuclear factor B in promoting double positive thymocyte apoptosis. J. Exp. T. Yokota. 1995. Cyclic AMP inhibits expression of the IL-2 gene through the Med. 189: 145–158. nuclear factor of activated T cells (NF-AT) site, and transfection of NF-AT 5. Voll, R. E., E. Jimi, R. J. Phillips, D. F. Barber, M. Rincon, A. C. Hayday, cDNAs abrogates the sensitivity of EL-4 cells to cyclic AMP. J. Immunol. 154: R. A. Flavell, and S. Ghosh. 2000. NF-B activation by the pre-T cell receptor 5255–5264. serves as a selective survival signal in T lymphocyte development. Immunity 13: 32. Schreiber, E., P. Matthias, M. M. Muller, and W. Schaffner. 1989. Rapid detec- 677–689. tion of octamer binding proteins with “mini-extracts” prepared from a small http://www.jimmunol.org/ 6. Senftleben, U., Z. Li, V. Baud, and M. Karin. 2001. IKKb is essential for pro- number of cells. Nucleic Acids Res. 17: 6419. tecting T cells from TNF␣-induced apoptosis. Immunity 14: 217–230. 33. Gautier, L., L. Cope, B. M. Bolsta, and R. A. Irizarry. 2004. Affy-analysis of 7. Fiorini, E., I. Schmitz, W. E. Marissen, S. L. Osborn, M. Touma, T. Sasada, Affymetrix GeneChip data at the probe level. Bioinformatics 20: 307–315. P. A. Reche, E. V. Tibaldi, R. E. Hussey, A. M. Kruisbeek, et al. 2002. Peptide- 34. Li, C., and H. W. Wong. 2001. Model-based analysis of oligonucleotide arrays: induced negative selection of thymocytes activates transcription of an NF-B expression index computation and outlier detection. Proc. Natl. Acad. Sci. USA inhibitor. Mol. Cell. 9: 637–648. 98: 31–36. 8. Kontgen, F., R. Grumont, A. Strasser, D. Metcalf, R. Li, D. Tarlinton, and 35. Chilson, O. P., and A. E. Kelly-Chilson. 1989. Mitogenic lectins bind to the S. Gerondakis. 1995. Mice lacking the c-rel proto-oncogene exhibit defects in antigen receptor on human lymphocytes. Eur. J. Immunol. 19: 389–396. lymphocyte proliferation, humoral immunity, and interleukin-2 expression. 36. Shibata, K., M. Imarai, G. M. van Bleek, S. Joyce, and S. G. Nathenson. 1992. Genes Dev. 9: 1965–1977. Vesicular stomatitis virus antigenic octapeptide N52–59 is anchored into the 9. Liou, H.-C., Z. Jin, J. Tumang, S. Andjelic, K. A. Smith, and M.-L. Liou. 1999. groove of the H-2Kb molecule by the side chains of three amino acids and the c-Rel is crucial for lymphocyte proliferation but dispensable for T cell effector main-chain atoms of the amino terminus. Proc. Natl. Acad. Sci. USA 89: by guest on October 2, 2021 function. Int. Immunol. 11: 361–371. 3135–3139. 10. Hwang, E. S., J.-H. Hong, and L. H. Glimcher. 2005. IL-2 production in devel- 37. Lanier, L. L. 2005. NK cell recognition. Annu. Rev. Immunol. 23: 225–274. oping Th1 cells is regulated by heterodimerization of RelA and T-bet and requires 38. Durand, D. B., J. P. Shaw, M. R. Bush, R. E. Replogle, R. Belagaje, and T-bet serine residue 508. J. Exp. Med. 202: 1289–1300. G. R. Crabtree. 1988. Characterization of antigen receptor response elements 11. Verma, I. M., J. K. Stevenson, E. M. Schwarz, D. Van Antwerp, and within the interleukin-2 enhancer. Mol. Cell Biol. 8: 1715–1724. S. Miyamoto. 1995. Rel/NF- B/I B family: intimate tales of association and 39. Novak, T. J., P. M. White, and E. V. Rothenberg. 1990. Regulatory anatomy of dissociation. Genes Dev. 9: 2723–2735. the murine interleukin-2 gene. Nucleic Acids Res. 18: 4523–4533. 12. Kunsch, C., S. M. Ruben, and C. A. Rosen. 1992. Selection of optimal kB/Rel 40. Jain, J., C. Loh, and A. Rao. 1995. Transcriptional regulation of the IL-2 gene. DNA-binding motifs: interaction of both subunits of NF- B with DNA is re- Curr. Opin. Immunol. 7: 333–342. quired for transcriptional activation. Mol. Cell Biol. 12: 4412–4421. 41. Powell, J. D., J. A. Ragheb, S. Kitagawa-Sakakida, and R. H. Schwartz. 1998. 13. Lin, R., D. Gewert, and J. Hiscott. 1995. Differential transcriptional activation in Molecular regulation of interleukin-2 expression by CD28 co-stimulation and vitro by NF- B/Rel proteins. J. Biol. Chem. 270: 3123–3131. anergy. Immunol. Rev. 165: 287–300. 14. Udalova, I. A., R. Mott, D. Field, and D. Kwiatkowski. 2002. Quantitative predi- 42. Doi, T. S., T. Takahashi, O. Taguchi, T. Azuma, and Y. Obata. 1997. NF-B cition of NF-B DNA-protein interactions. Proc. Natl. Acad. Sci. USA 99: RelA-deficient lymphocytes: normal development of T cells and B cells, impaired 8167–8172. production of IgA and IgG1 and reduced proliferative responses. J. Exp. Med. 15. Sanjabi, S., K. J. Williams, S. Saccani, L. Zhou, A. Hoffmann, G. Ghosh, 185: 953–961. S. Gerondakis, G. Natoli, and S. T. Smale. 2005. A c-Rel subdomain responsible for enhanced DNA-binding affinity and selective gene activation. Genes Dev. 19: 43. Sen, R., and D. Baltimore. 1986. Multiple nuclear factors interact with the im- Cell 2138–2151. munoglobulin enhancer sequences. 46: 705–716. 16. Saccani, S., S. Pantano, and G. Natoli. 2003. Modulation of NF-B activity by 44. Fraser, J. D., B. A. Irving, G. R. Crabtree, and A. Weiss. 1991. Regulation of exchange of dimers. Mol. Cell. 11: 1563–1574. interleukin-2 gene enhancer activity by the T-cell accessory molecule CD28. 17. Ghosh, S., M. J. May, and E. B. Kopp. 1998. NF-B and Rel proteins: evolu- Science 251: 313–316. tionarily conserved mediators of immune responses. Annu. Rev. Immunol. 16: 45. Shapiro, V. S., K. E. Truitt, J. Imboden, and A. Weiss. 1997. CD28 mediates 225–260. transcriptional upregulation of the interleukin-2 (IL-2) promoter through a com- 18. Kitamura, H., K. Kanehira, K. Okita, M. Morimatsu, and M. Saito. 2000. MAIL, posite element containing the CD28RE and NF-IL-2B AP-1 sites. Mol. Cell Biol. a novel nuclear IB protein that potentiates LPS-induced IL-6 production. FEBS 17: 4051–4058. Lett. 485: 53–56. 46. Garrity, P. A., D. Chen, E. V. Rothenberg, and B. J. Wold. 1994. Interleukin-2 19. Haruta, H., A. Kato, and K. Todokoro. 2001. Isolation of a novel interleukin-1- transcription is regulated in vivo at the level of coordinated binding of both inducible nuclear protein bearing ankyrin-repeat motifs. J. Biol. Chem. 276: constitutive and regulated factors. Mol. Cell Biol. 14: 2159–2169. 12485–12488. 47. Park, G. T., S. E. Lim, S.-I. Jang, and M. I. Morasso. 2002. Suprabasin, a novel 20. Yamazaki, S., T. Muta, and K. Takeshige. 2001. A novel IB protein, IB-, epidermal differentiation marker and potential cornified envelope precursor. induced by proinflammatory stimuli, negatively regulates nuclear factor-Binthe J. Biol. Chem. 277: 45195–45202. nuclei. J. Biol. Chem. 276: 27657–27662. 48. Yamaji, K., H. Ikegami, T. Fujisawa, S. Noso, K. Nojima, N. Babaya, 21. Yamamoto, M., S. Yamazaki, S. Uematsu, S. Sato, H. Hemmi, K. Hoshino, M. Itoi-Babaya, S. Makino, T. Sakamoto, and T. Ogihara. 2005. Evidence for T. Kaisho, H. Kuwata, O. Takeuchi, K. Takeshige, et al. 2004. Regulation of CD101 but not Fcgr1 as candidate for type 1 diabetes locus, Idd10. BBRC 331: Toll/IL-1-receptor-mediated gene expression by the inducible nuclear protein 536–542. IB. Nature 430: 218–222. 49. Soares, L. R. B., L. Tsavaler, A. Rivas, and E. G. Engleman. 1998. V7 (CD101) 22. Kuwata, H., N. Watanabe, H. Miyoshik, M. Yamamoto, T. Kaisho, K. Takeda, ligation inhibits TCR/CD3-induced IL-2 production by blocking Ca2ϩ flux and and S. Akira. 2003. IL-10-inducible Bcl-3 negatively regulates LPS-induced nuclear factor of activated T cell nuclear translocation. J. Immunol. 161: TNF-␣ production in macrophages. Blood 102: 4123–4129. 209–217. 1692 IBNS REGULATION OF T CELL CYTOKINES
50. Gu, X., A. Laouar, J. Wan, M. Daheshia, J. Lieberman, W. M. Yokoyama, 56. Dzhagalov, I., V. Giguere, and Y.-W. He. 2004. Lymphocyte development and H. R. Katz, and N. Manjunath. 2003. The gp49B1 inhibitory receptor regulates function in the absence of retinoic acid-related orphan receptor a. J. Immunol. the IFN-␥ responses of T cells and NK cells. J. Immunol. 170: 4095–4101. 173: 2952–2959. 51. Colotta, F., F. Re, M. Muzio, R. Bertini, N. Polentarutti, M. Sironi, J. G. Giri, 57. Rao, S., S. Gerondakis, D. Woltring, and M. F. Shannon. 2003. c-Rel is required S. K. Dower, J. E. Sims, and A. Mantovani. 1993. Interleukin-1 type II receptor: for chromatin remodeling across the IL-2 gene promoter. J. Immunol. 170: a decoy target for IL-1 that is regulated by IL-4. Science 261: 472–475. 3724–3731. 52. Xin, X., O. Varlamov, R. Day, W. Dong, M. M. Bridgett, E. H. Leiter, and 58. Takahashi, I., H. Miyaji, T. Yoshida, S. Sato, and T. Mizukami. 1996. Selective L. D. Fricker. 1997. Cloning and sequence analysis of cDNA encoding rat car- inhibition of IL-2 gene expression by trichostatin A, a potent inhibitor of mam- boxypeptidase D. DNA Cell Biol. 16: 897–909. malian histone deacetylase. J. Antibiot. 49: 453–457. 59. Chen, L. F., and W. C. Greene. 2003. Regulation of distinct biological activities 53. Veillette, A., M.-E. Cruz-Munoz, and M.-C. Zhong. 2006. SLAM family recep- of the NF-B transcription factor complex by acetylation. J. Mol. Med. 81: tors and SAP-related adaptors: matters arising. Trends Immunol. 27: 228–234. 549–557. 54. Kanno, Y., B. Z. Levi, T. Tamura, and K. Ozato. 2005. Immune cell-specific 60. Himes, S. R., L. S. Coles, R. Reeves, and M. F. Shannon. 1996. High mobility amplification of interferon signaling by the IRF-4/8-PU.1 complex. J. Interferon group protein I(Y) is required for function and for c-Rel binding to CD28 re- Cytokine Res. 25: 770–779. sponse elements within the GM-CSF and IL-2 promoters. Immunity 5: 479–489. 55. Lange, U. C., M. Saitou, P. S. Western, S. C. Barton, and M. A. Surani. 2003. The 61. Himes, S. R., R. Reeves, J. Attema, M. Nissen, Y. Li, and M. F. Shannon. 2000. fragilis interferon-inducible gene family of transmembrane proteins is associated The role of high-mobility group I(Y) proteins in expression of IL-2 and T cell with germ cell specification in mice. BMC Dev. Biol. 3: 1–11. proliferation. J. Immunol. 164: 3157–3168. Downloaded from http://www.jimmunol.org/ by guest on October 2, 2021