The development and activation of MHC class II (MHC-II)–restricted CD4+ T cells are distinct immunological processes that are strictly MHC-II–dependent. To address their relative dependence on MHC-II, we established a novel ENU-induced mutant mouse on the C57BL/6 background, named I-A12%, with ∼8-fold reduced I-A expression on the surface of B cells, dendritic cells, cortical thymic epithelial cells, and medullary thymic epithelial cells. I-A100% and I-A12% mice are highly similar with respect to the numbers of double-positive thymocytes, CD4+CD8 T cells, regulatory T cells, CD4+ T cell marker expression, lifespan, and Th/regulatory T cell function. Despite the demonstration of functional intrathymic negative selection in I-A12% mice, transfer of I-A12% CD25CD4+ T cells into RAG-knockout hosts revealed increased autoaggression activity against the liver. Compared to I-A100% mice, infection of I-A12% mice with graded doses of Listeria monotcytogenes or influenza virus revealed comparable and significantly reduced generation of Ag-specific CD4+ T cells at high and low infection doses, respectively. A significantly weakened Ag-specific recall cytokine production response was also found for I-A12% mice previously infected with a relative low dose of L. monocytogenes. CD44hiCD4+ T cells from I-A100% and I-A12% mice previously infected with a relatively high L. monocytogenes dose displayed highly similar Ag-specific multicytokine production profiles. In contrast, polyclonal activation of endogenous memory-like I-A12% CD44hiCD4+ T cells revealed highly elevated production of multiple cytokines. Our results demonstrate that there exist distinct thresholds for different MHC-II–dependent immunological processes. The I-A12% mutant mouse model we describe in the present study is a valuable tool for investigations on the quantitative cause–effect relationship in MHC-II–dependent normal and autoimmune responses.

In the total absence of MHC class II (MHC-II), CD4+ T cell development is completely halted (1, 2). MHC-II molecules are made up of α- and β-chains and in the mouse, and there are two MHC-II heterodimers, I-Aα:I-Aβ (I-A) and I-Eα:I-Eβ (I-E). I-E is nonfunctional in the common C57BL/6 (B6) inbred mouse strain due to a mutation in I-Eα, leaving I-A as the only functional MHC-II. I-A–expressing cells belonging to different lineages perform distinct and indispensable functions. Inside the thymus where T cell development takes place, I-A+ cortical thymic epithelial cells (cTECs) mediate positive selection of I-A–restricted CD4+ T cells, and I-A+ medullary TECs (mTECs) and dendritic cells (DCs) mediate negative selection of autoreactive CD4+ T cells (3). In the periphery, CD4+ T cell immune responses are initiated through TCR-mediated specific recognition of antigenic peptides bound to I-A on the surface of DCs, followed by activation and expression of effector function. I-A expression is under the regulation of the master regulator CIITA and that the amount of expressed I-A is positively correlated with that of CIITA (4, 5). The CIITA gene contains three functional promoters (pI, pIII, and pIV) that are employed by cells of myeloid, lymphoid, and nonhematopoietic (such as TECs) origins, respectively, and the genomic organization of the CIITA gene is evolutionarily conserved (5). Although I-A is clearly important for T cell development, the reason behind the conserved high level of I-A expression in cTECs, mTECs, and DCs is not fully understood. Specifically, the relative quantity of I-A required for positive selection, negative selection, and induction of an antigenic response in the periphery has not been carefully determined. In this study, through the establishment of a novel mutant mouse, named I-A12%, with a generalized reduced I-A expression on cTECs, mTECs, DCs, and B cells, we present experimental results that support distinct I-A thresholds for different I-A–dependent immunological processes.

All experiments involving mice adhered to active protocols approved by Institutional Animal Care and Use Committee of Academia Sinica. Unless otherwise indicated, all mice used were 4–12 wk of age and were bred and housed under specific pathogen-free conditions in the Animal Facility, Institute of Molecular Biology, Academia Sinica. B6, BALB/cBy, OVA transgenic (Tg; catalog no. 005145), OT-II TCR Tg (004194), TCRβ-knockout (KO; 002118) (6), RAG1-KO (002216), and C3H/HeJ breeders were originally obtained from The Jackson Laboratory. I-Aβ–KO (1) and AND TCR Tg mice (7) mice were originally provided by Dr. B.J. Fowlkes (Laboratory of Immunology, National Institute of Allergy and Infectious Diseases, National Institutes of Health) and Dr. S. Hedrick (University of California San Diego, San Diego, CA), respectively. The B6 histocompatible B6.TL (H-2bThy1aCD8a) congenic strain was as previously described (8).

The I-A12% mutant mouse described in the present study was an ENU-induced mutant that was generated by treating B6 male mice by ENU, followed by a three-generation breeding scheme to generate G3 mice (9). Linkage of the causative mutant gene to Ch17 was established by haplotype interval analysis of affected (G3 × BALB/c) F2 mice (10). Through screening of ENU G3 mice, a mutant (I-A12%) with highly reduced I-A (MHC-II) expression on PBLs was identified. The conserved A located in the branch site of intron 2 of the I-Aα gene was mutated to T (see Fig. 1D) in I-A12% mice, causing pre-mRNA splicing errors and ∼8-fold reduced expression of wild-type I-Aα mRNA and surface-bound I-A.

FIGURE 1.

Mutation in I-Aα is responsible for reduced I-A expression in I-A12% mutant mice. (A) PBLs from B6 and ENU mutant mice, named I-A12%, were stained as described in 2Materials and Methods, followed by flow cytometric analysis. I-A expression by B cells is shown. (B) An affected I-A12% G3 male mouse was mated with BALB/c females to generate F1 mice, which were allowed to intercross to generate F2 mice. Haplotype interval analysis of affected (low I-A expression) F2 mice (n = 8) revealed a strong linkage to the marker D17mit46 (17-1) on chromosome 17 at the 25.5 Mbp position, based on the least number of nonrecombinant haplotypes and the highest χ2 score (marked by an asterisk). (C) Affected G3 mice were crossed with B6 mice to generate F1 offspring. F1 mice were intercrossed to generate F2 mice, 22% of which expressed highly reduced I-A. Further mating between affected F2 mice resulted in offspring that were 100% affected. Numbers shown indicate the total number of mice analyzed. (D) DNA sequencing of the I-Aα gene revealed an A to T point mutation in the splicing branch site of I-Aα intron 2. (E) DNA sequence covering exon 2 (3′ end) to exon 3 (5′ end) of the I-Aα gene is shown. Sequences for exons 2 and 3 are capitalized and boxed. Intron 2 spans 431 bp and is numbered 1–431 and shown in lowercase letters. The conserved A at position 408 of the putative CTCAT branch site (underlined, position 405–409) was mutated to T in I-A12% mice. (F) Full-length I-Aα mRNA of I-A100% and I-A12% spleen follicular B cells was analyzed by RT-PCR. PCR products of I-A100% mice yielded one major band designated I-A-Wt. There were three distinct-sized PCR products for I-A12%: one identical to I-A-Wt, and the other two PCR products marked by I-A-1 and I-A-2. For I-A-1 mRNA, an alternative splice acceptor site at (position 198 of intron 2, underlined) was used, resulting in out-of-frame translation and premature termination upon encountering TGA (position 215–217). For I-A-2 mRNA, intron 2 was not spliced out, resulting in translation of 5′ intron 2 sequences until a premature termination TGA (position 129–131 of intron 2) was encountered. All data shown are representative of three independently performed experiments.

FIGURE 1.

Mutation in I-Aα is responsible for reduced I-A expression in I-A12% mutant mice. (A) PBLs from B6 and ENU mutant mice, named I-A12%, were stained as described in 2Materials and Methods, followed by flow cytometric analysis. I-A expression by B cells is shown. (B) An affected I-A12% G3 male mouse was mated with BALB/c females to generate F1 mice, which were allowed to intercross to generate F2 mice. Haplotype interval analysis of affected (low I-A expression) F2 mice (n = 8) revealed a strong linkage to the marker D17mit46 (17-1) on chromosome 17 at the 25.5 Mbp position, based on the least number of nonrecombinant haplotypes and the highest χ2 score (marked by an asterisk). (C) Affected G3 mice were crossed with B6 mice to generate F1 offspring. F1 mice were intercrossed to generate F2 mice, 22% of which expressed highly reduced I-A. Further mating between affected F2 mice resulted in offspring that were 100% affected. Numbers shown indicate the total number of mice analyzed. (D) DNA sequencing of the I-Aα gene revealed an A to T point mutation in the splicing branch site of I-Aα intron 2. (E) DNA sequence covering exon 2 (3′ end) to exon 3 (5′ end) of the I-Aα gene is shown. Sequences for exons 2 and 3 are capitalized and boxed. Intron 2 spans 431 bp and is numbered 1–431 and shown in lowercase letters. The conserved A at position 408 of the putative CTCAT branch site (underlined, position 405–409) was mutated to T in I-A12% mice. (F) Full-length I-Aα mRNA of I-A100% and I-A12% spleen follicular B cells was analyzed by RT-PCR. PCR products of I-A100% mice yielded one major band designated I-A-Wt. There were three distinct-sized PCR products for I-A12%: one identical to I-A-Wt, and the other two PCR products marked by I-A-1 and I-A-2. For I-A-1 mRNA, an alternative splice acceptor site at (position 198 of intron 2, underlined) was used, resulting in out-of-frame translation and premature termination upon encountering TGA (position 215–217). For I-A-2 mRNA, intron 2 was not spliced out, resulting in translation of 5′ intron 2 sequences until a premature termination TGA (position 129–131 of intron 2) was encountered. All data shown are representative of three independently performed experiments.

Close modal

PBLs were stained by Alexa Fluor (A) 488–anti-I-A (clone Y3P) and A405–anti-B220 (clone RA3-6B2), and B cells (B220+) were analyzed for I-A expression by flow cytometry (LSR II; BD Biosciences). Spleen cells were stained by A488–anti-I-A, A647)–anti-CD11c (clone N418), A680–anti-CD11b (clone M1/70), and A405-anti-B220; I-A expression was analyzed by B cells (B220+), and DCs (CD11c+CD11b+) were analyzed by flow cytometry. To prepare cTECs and mTECs, thymi were removed and subjected to collagenase IV (Life Technologies) and DNAse I (Roche) treatment as previously described (11), and they were stained by either 1) A488–anti-I-A, allophycocyanin–anti-Epcam (BD Biosciences), A405–anti-CD45 (clone M1/9.3.4.HL.2), and PE–anti-BP1 (BD Biosciences) or A546–UEA-1 (Vector Laboratories), or 2) A488–anti-I-A, A647-CD11c, and A405–anti-CD45. I-A expression on cTECs (CD45Epcam+BP1+), mTECs (CD45Epcam+UEA1+), and DCs (CD45+CD11c+) was determined by flow cytometric analysis. For T cell analysis, thymus and spleen cells were stained with combinations of fluorochrome-conjugated mAbs from the following: PE–anti-CD4 (clone GK1.5), A405–anti-CD8 (clone 53-6.7), A488–anti-TCRβ (clone H57-597), A647–anti-CD69 (clone H1-2F3), A488–anti-CD25 (clone 7D4), A647–anti-CD44 (clone IM7), A647–anti-CD62L (clone Mel14), A647–anti-Ly6C (clone S12C1), FITC–anti-CD122 (BD Biosciences), Qa2 (clone 1-9-9), CD5 (clone 53-7.3), A647–anti-Vβ2 (clone B20.6), A647–anti-Vβ3 (clone KJ25), FITC-Vβ4 (BD Biosciences), FITC-Vβ5 (clone MR9-4), A647–anti-Vβ6 (clone RR4-7), A647-Vβ7 (clone TR310), A647–anti-Vβ8 (clone F23.1), FITC–anti-Vβ9 (BD Biosciences), FITC–anti-Vβ10 (BD Biosciences), FITC–anti-Vβ11 (clone RR3-15), FITC–anti-Vβ12 (BD Biosciences), FITC–anti-Vβ13 (BD Biosciences), FITC–anti-Vβ14 (clone 14-2), A647–anti-Vα2 (clone B20.1), A647–anti-Vα3.2 (clone RR3-16), FITC-anti-Vα8 (clone KT50), A647-anti-Vα11 (clone RR8-1), and A488–anti-Thy1.2 (clone 30H12). To determine the expression of surface Ags, live cells were always stained by indicated fluorochrome-conjugated Abs in the presence of unconjugated anti-FcR (clone 2.4G2) mAb to inhibit FcR-mediating binding, as well as propidium iodide (PI). Dead cells that have incorporated PI were gated out and only live cells were analyzed. The various cell types being analyzed were additionally gated by their known characteristic forward light scatter (FSC)/side light scatter (SSC) properties: low FSC/SSC for lymphocytes, high FSC/SSC for cTECs, mTECs, and DCs.

Intracellular Foxp3 was detected by A647–anti-Foxp3 (clone FJK-16s, eBioscience) after cell fixation and permeabilization using a Foxp3 staining buffer kit (eBioscience). FSC/SSC gating characteristic of Foxp3+ small lymphocytes was applied in the analysis Foxp3 expression. For active caspase-3 detection, thymocytes were first stained with fixable viability dye (eFluor 780; eBioscience) per the manufacturer’s instructions, and then by A488–anti-CD69, PE–anti-CD4, and A405–anti-CD8, fixed, and permeabilized (fixation/permeabilization solution kit; BD Biosciences), followed by staining with A647–anti-cleaved caspase-3 (Cell Signaling Technology). FSC/SSC gating characteristic of resting lymphocytes, as well as viable cell gating (fixable viability dye eFluor 780+), was applied in the analysis of active caspase-3 expression in live cells. Flow cytometric data were analyzed by FlowJo software (Tree Star).

For imaging flow cytometric analysis, cells were stained by PE–anti-CD4 and A488–anti-CD8, followed by Hoechst staining for 90 min at 37°C, and then analyzed by image flow cytometry (ImageStream Mark II, equipped with 405-, 488-, 642-, and 785-nm darkfield lasers, ×40 objective, Amnis). Acquired multispectral data were analyzed by the Amnis IDEAS software package. Cells analyzed were first gated by “Area” (size feature) versus “Aspect Ratio” (shape feature) parameters to exclude cell debris, and the “Gradient RMS” (texture feature) parameter was then used to determine the best focus. Cells with fragmented nuclei were empirically found to be enriched within low aspect ratio intensity (defined by the minor axis intensity divided by the major axis intensity) and low circularity morphology (defined by the average distance of the object boundary from its center divided by the variation of this distance to differentiate between rounded objects with smooth boundary to less regular objects).

Spleen B cells, DCs, cTECs, mTECs, and thymus DCs were isolated by cell sorting as indicated. RNA extraction, DNase I treatment, and reverse transcription were performed as described (12). Expression of I-Aα (exons 2–3) mRNA was analyzed by real-time quantitative PCR and normalized against I-Aβ as previously described (12). Full-length I-Aα mRNA (796 bp) was analyzed for spleen B cells by conventional PCR. For quantitation, region-specific segments for I-Aα and I-Aβ mRNAs were subjected to real-time PCR. Primers used were: for conventional I-Aα (full-length) PCR, forward, 5′-CCAGGATGCCGCGCAGCAGA-3′, reverse, 5′-GTCACACCCTGGAAAGGAAG-3′; for real-time PCR, I-Aα (150 bp), forward, 5′-CAGCTACCAATGAGGCTCCTCAAG-3′ (spanning exons 2 and 3), reverse, 5′-CGTCTGCGACTGACTTGCTA-3′; and for real-time I-Aβ (115 bp), forward, 5′-CACTCTGGTCTGCTCAGTGACAG-3′, reverse, 5′-CCCATTCCTAATAAGCTGTGTGG-3′. The conditions of real-time PCR were: 95°C (10 s), 58°C (5 s), and 72°C (6 and 5 s for I-Aα [exons 2–3] and I-Aβ, respectively). For full-length I-Aα, the conditions were: 35 cycles of 94°C (30 s), 56°C (45 s), and 72°C (40 s).

Cell turnover in mice was studied after implanting 5′-ethynyl-2′-deoxyuridine (EdU; Invitrogen)–filled mini-osmotic pumps (Alzet no. 2001, 2.8 mg of EdU in 200 μl per pump) s.c. along the midline of the back. The implanted pump allowed the release of EdU at a constant rate during a period of 7 d, at which time the osmotic pump was surgically removed. The turnover of EdU-labeled cells was studies during a period of 0, 3, 7, 14, and 21 d after EdU pump removal. Thymocytes and spleen cells were separately stained with A647–anti-TCRβ, A405–anti-CD8, allophycocyanin-Cy7–anti-CD4, and A488-azide to detect EdU according to the manufacturer’s instructions (Invitrogen). On the basis of their marker expression, indicated subsets (thymic double-negative [DN], CD4CD8; double-positive [DP], CD4+CD8+; CD4 single-positive (SP), CD4+CD8TCRβ+; CD8SP, CD4CD8+TCRβ+; and spleen CD4SP, CD4+CD8TCRβ+ T cells) were analyzed for changes of percentage EdU+ cells over time.

MLR cultures were set up by mixing responding CD4+ T cells (2 × 105 cells per 100 μl per well) isolated either by panning or cell sorting of NK1TCRδCD4+CD8 T cells (2 × 105 cells per 100 μl per well) from indicated mice as described previously (13) and syngeneic or allogeneic CD11c+ stimulator cells (1.2–1.5 × 105 cells per well) as indicated, or syngeneic T cell–depleted spleen cells at indicated cell numbers, with or without added anti-CD4 (clone GK1.5, 1 μg/ml). Cell proliferation was measured on day 5 by a 6-h [3H]thymidine pulse. The total number of activated CD4+ T cells per culture was determined with the aid of absolute counting beads. Day 5 MLR culture wells were stained individually with FITC–anti-CD25, A647–anti-TCRβ, A405–anti-CD4, and PI, with CountBright absolute counting beads (20,000 beads per well; Thermo Fisher Scientific) added immediately prior to flow cytometric analysis. Absolute numbers of activated CD4+ T cells per culture were calculated as (live [PI]TCRβ+CD4hiCD25+ number ÷ bead number) × 20,000. B6 congenic B6.TL (H2bThy1aCD8a) host mice were irradiated (750 rad), followed by reconstitution with bone marrow cells (3 × 106 cells per mouse) from I-A100%, I-A12%, and I-A0% mice. At 2 mo after reconstitution, donor-origin CD4+ T cells (Thy1.2+CD4+CD8) were isolated by cell sorting and used as responder cells in an MLR in which either allogeneic C3H or syngeneic B6 CD11c+ DCs were used as stimulator cells. Mitogenic activation cultures were set up with CD4+ T cells (2 × 104 cells per well) and indicated concentrations of anti-CD3 mAb (clone 500A.A2) presented by B6 B cell blasts (105 cells per well) as described previously (12). Cell proliferation was measured on day 3 by a 6-h [3H]thymidine pulse.

Sorted spleen Vα2+Vβ5+CD4+8 cells (2 × 104 cells per well) from OT-II TCR Tg mice on I-A100% and/or I-A12% background were stimulated by OVA323–339 peptide or OVA protein presented by B6 B cell blasts (105 cells per well). Cell proliferation was measured on day 3 by a 6-h [3H]thymidine pulse. Day 3 culture supernatants collected from OVA323–339-stimulated OT-II T cells or from plate-bound 10 μg/ml anti-CD3 (clone 500A.A2)/anti-CD28 (clone 37.51)–stimulated CD4+ T cells (105 cells per well) were assayed for IL-2 using CTLL indicator cells as previously described (8). In other experiments involving IL-2 assay, a cytokine bead array (CBA) assay (see below) rather than a bioassay was used. The amount of IL-2 determined by the bioassay and the CBA assay are highly similar and we use these assays interchangeably. Electronic cell sorting was performed using a cell sorter (BD FACSAria II) equipped with 407-, 488-, and 633-nm lasers. The purity of sorted cells was always checked by reanalysis of the sorted cell subpopulations. The purity of the sorted cells was usually >99%, but was always >98%.

To assess T-dependent Ab responses, TCRβ-KO mice were adoptively transferred (i.v.) with CD4+ T cells (5 × 106 cells per mouse) from either I-A100% or I-A12% mice 1 d prior to immunization (i.p.) with 100 μg of 4-hydroxy-3-nitrophenylacetyl (NP) hapten–conjugated chicken γ-globulin (NP-CCG; Biosearch Technologies) that had been admixed 1:1 with Imject Alum (Thermo Fisher Scientific). Sera were harvested before immunization and on days 16 and 32 after immunization. High-affinity anti-NP Abs were assessed by applying serum to ELISA plates coated with NP3-BSA, followed by biotin-conjugated anti-IgG1 (clone A85-1) or anti-IgG2a (clone R19-15) Ab (BD Biosciences). Biotin groups were then detected by HRP-streptavidin (Pierce), followed by addition of HRP substrate ABTS (Sigma-Aldrich) and absorbance reading at 415 nm. IgG1 and IgG2a standard curves were set up by coating microtiter wells with anti-IgG1 (clone A85-3) and anti-IgG2a capture mAb (clone R11-89; BD Biosciences), respectively. Graded amounts of known IgG1 (clone MR9-4) and IgG2a (clone 14-4-4) mAbs were then added to wells coated with anti-IgG1 (clone A85-3) and IgG2a (clone R11-89), respectively. The captured IgG1 and IgG2a were then detected by biotin-conjugated anti-IgG1 and anti-IgG2a mAbs, respectively, followed by addition of HRP substrate ABTS (Sigma-Aldrich), color development, and absorbance reading (415 nm). Half-maximal OD (415 nm) absorbance was arbitrarily defined as 1 U/ml. All test serum samples were diluted such that their OD (415 nm) readings were in the titratable range of the IgG1/IgG2a standard curve and subjected to extrapolation to derive relative amounts of NP-reactive IgG1 and IgG2a, expressed in units per milliliter amounts.

To induce colitis and liver damage, I-A100% and I-A12% spleen cells were stained with FITC–anti-CD45RB (clone 23G2), PE–anti-CD4, and A647–anti-CD25, followed by cell sorting to isolate CD4+CD25CD45RBhi and CD4+CD25+CD45RBlo subsets. RAG1-KO mice were adoptively transferred (i.p.) with sorted CD4+CD25CD45RBhi cells (5 × 105 cells per host) along with or without sorted CD4+CD25+CD45RBlo cells (105 cells per host) as indicated. To assess colitis development, body weight of the RAG1-KO hosts was recorded weekly. When 20% of their original body weight had been lost, RAG1-KO hosts were sacrificed and their colons (cecum to anus) collected and weighed.

Liver damage in RAG1-KO hosts that had received adoptive transfer of I-A100% or I-A12% CD4+CD25CD45RBhi T cells (5 × 105 cells per host) was assessed at 7 wk after adoptive transfer: 1) serum samples were assayed for alanine aminotransferase (ALT) and aspartate aminotransferase (AST) (Cobas Integra 400 Plus; Roche); and 2) liver samples were fixed (10% neutral formalin), embedded in paraffin, sectioned (5 μm thickness), H&E stained, and examined by light microscopy. Liver damage was scored using the following scheme: cell infiltration (0, none; 1, mild; 2, moderate; 3, severe) and necrotic lesions (0, none; 1, mild; 2, moderate; 3, severe). The combined maximal score is 6, with 0 being the minimal score.

Indicated mice were challenged with L. monocytogenes strain 10403S (StrepR) as previously described (14). I-A100% and I-A12% mice were injected (i.v.) with a sublethal dose of L. monocytogenes as indicated, sacrificed 5 wk later, and their spleen CD4+ T cells were isolated by cell sorting and stimulated (2.5 × 105 cells per 100 μl per well) by addition of 1 μM final concentration of the dominant listeriolysin O (LLO)190 or the subdominant LLO318 antigenic peptide (15), which were presented by I-A+ B6 B cell blasts (2.5 × 105 cells per well). IL-2 and IFN-γ released into day 2 culture supernatants were detected by CBA (BD Biosciences). To detect MHC-II (I-Ab)–restricted L. monocytogenes–specific CD4+ T cells that recognize the dominant LLO190–201 peptide, spleen cells from mice that had previously been infected with L. monocytogenes were stained with A488–anti-CD8, A647–anti-CD4, and Brilliant Violet (BV)421–I-Ab–LLO190–201 tetramer (National Institutes of Health Tetramer Core Facility), followed by flow cytometric analysis. For negative control, BV421–I-Ab–human CLIP87–101 tetramer (National Institutes of Health Tetramer Core Facility) was used instead of the BV421–I-Ab–LLO190–201 tetramer.

Influenza A/WSN/33 (H1N1) virus was generated using the eight-plasmid cotransfection system (courtesy of Dr. Robert G. Webster, University of Tennessee) as previously described (16). Freshly prepared influenza virus was used for all infection experiments. Fresh influenza virus was made by inoculating cultures of mixed 293T and BHK cells with a thawed influenza virus frozen stock. Titers of virus so grown were determined on a lawn of Madin–Darby canine kidney cells (provided by Dr. Wen Chang, Academia Sinica). Mice were infected with influenza virus intranasally at indicated doses within 2 d of PFU determination. To detect MHC-II (I-Ab)–restricted influenza virus-specific CD4+ T cells that recognize NP311–325, spleen cells from mice that had been previously infected with influenza virus were stained with A488–anti-CD8, A647–anti-CD4, and BV421–I-Ab–NP311–325 tetramer (National Institutes of Health Tetramer Core Facility), followed by flow cytometric analysis. For negative control, BV421–I-Ab–human CLIP87–101 tetramer (National Institutes of Health Tetramer Core Facility) was used instead of the BV421–I-Ab–NP311–325.

Spleen cells were stained with A488–anti-NK1 (clone PK136), A488–anti-TCRδ (clone GL3), A647–anti-CD44, PE–anti-CD4, and A405–anti-CD8. As TCR-γδ T cells and NKT cells are known to rapidly produce cytokines upon TCR stimulation, we employed a sorting strategy that eliminated TCR-γδ and NKT cells. Sort gates for naive and memory CD4+ T cells were CD44loNK1TCRδCD4+CD8 and CD44hiNK1TCRδCD4+CD8, respectively. Sorted T cells (105 cells per well) were stimulated with plate-bound anti-CD3/CD28 (10 μg/ml), and IL-2, IFN-γ, TNF-α, IL-4, IL-5, IL-6, IL-10, IL-13, and IL-17 released into day 1 and 2 culture supernatants were detected by CBA according to the manufacturer’s instructions (BD Biosciences). Spleen CD44hiCD4+CD8 T cells from mice infected with 5000 CFU of L. monocytogenes 5 wk previously were sorted and stimulated (105 cells per 100 μl per well) by LLO190 peptides (1 μM) presented by B6 B cell blasts (2 × 105 cells per well). Cytokines released into day 2 culture supernatants were assayed by CBA.

For most results, a two-tailed Student t test was used to calculate statistical difference between indicated experimental groups. In situations where experimental results assumed a non-normal distribution pattern, statistical difference was analyzed by an F test for unequal variance.

ENU-mutagenized G3 mice were screened for altered immune cell marker expression on PBLs by multiparameter flow cytometry. One ENU mutant, named I-A12%, displayed ∼8-fold reduced surface I-A expression on peripheral B cells (Fig. 1A). To identify the causative mutant gene, F2 mice were generated by mating affected I-A12% male mice (B6 background) to BALB/c female mice. Interval haplotype analysis of eight affected F2 mice showed linkage to chromosome 17 (Fig. 1B). F2 mice were also generated by mating affected I-A12% male mice to B6 female mice (Fig. 1C). Among the 136 F2 offspring analyzed, 30 showed the originally identified ∼8-fold reduced I-A expression, a phenotype most likely caused by a single mutant somatic gene, as the affected ratio of 22.1% approximated the expected 1:4 ratio for simple Mendelian inheritance. Based on chromosome 17 linkage, I-Aα and I-Aβ were considered the most likely candidate mutant genes. DNA sequencing revealed an A to T mutation (at position 408) in the branch site (CTCAT) of the 431-bp intron 2 of the I-Aα gene (Fig. 1D, 1E); no mutation was found for I-Aβ. As the conserved A of the branch site plays a critical role in splicing, likely altered splicing was next examined. A dominant I-Aα mRNA isoform was seen for B6, whereas two additional forms, I-Aα-1 and I-Aα-2, were seen for I-A12% (Fig. 1F). I-Aα-1 mRNA was formed by using the wild-type splice donor site (exon 2) and the TTCAT (position 166–170) splicing acceptor site located in intron 2, resulting in premature termination (Fig. 1E). I-Aα-2 mRNA was formed by reading through intron 2 without splicing, also resulting in premature termination. Of the three I-Aα mRNA species found in I-A12% mice, only one is functional.

More detailed comparison of surface I-A expression was examined for spleen B cells and DCs, and for thymic cTECs, mTECs, and DCs (Fig. 2A). Mean fluorescence intensity (MFI) values were used as an estimate of I-A abundance. MFI of B6 spleen B cells, spleen DCs, cTECs, mTECs, and thymic DCs were 7.8-, 9.1-, 7.0-, 7.1-, and 9.0-fold those of mutant I-A12% mice, respectively. The expression of I-Aα mRNA was also determined for spleen B cells, spleen DCs, cTECs, mTECs, and thymic DCs. All B6 cell subsets examined expressed I-Aα mRNA at levels between 7.7- and 9.1-fold those of I-A12% mice (Fig. 2B). I-A12% is therefore a mouse model with a generalized ∼8-fold less I-A expression. For simplicity and clarity, we named the mutant mouse I-A12%. To be consistent, normal B6 mice are referred to as I-A100% throughout.

FIGURE 2.

I-A12% is a mutant mouse model with a generalized reduced I-A expression in all lineages of I-A–bearing cells. (A) B6 and I-A12% spleen cells were stained as described in 2Materials and Methods, followed by flow cytometric analysis. I-A expression by B cells and DCs is shown. B6 and I-A12% cTECs, mTECs, and DCs were prepared and stained as described in 2Materials and Methods. Histograms of I-A expression by cTECs, mTECs, and DCs are shown. MFI values for spleen B cells (with I-A100% and I-A12% denoting B6 and I-A12%, respectively) were: I-A100%, 5295; I-A12%, 677; for spleen DCs: I-A100%, 5910; I-A12%, 652; for cTECs: I-A100%, 3544; I-A12%, 506; for mTECs: I-A100%, 3659, I-A12%, 516; and for thymic DCs: I-A100%, 4663, I-A12%, 517. (B) I-A100% and I-A12% spleen cells and thymocytes were stained as in (A); indicated subsets were obtained by cell sorting and then analyzed for I-Aα and I-Aβ mRNA expression by real-time PCR (forward [F] and reverse [R] primer positions as indicated). To avoid detection of I-A-1 and I-A-2, the forward primer spanned sequences from both exon 2 and exon 3. Data shown are representative of two independently performed experiments. *p ≤ 0.05, **p ≤ 0.01.

FIGURE 2.

I-A12% is a mutant mouse model with a generalized reduced I-A expression in all lineages of I-A–bearing cells. (A) B6 and I-A12% spleen cells were stained as described in 2Materials and Methods, followed by flow cytometric analysis. I-A expression by B cells and DCs is shown. B6 and I-A12% cTECs, mTECs, and DCs were prepared and stained as described in 2Materials and Methods. Histograms of I-A expression by cTECs, mTECs, and DCs are shown. MFI values for spleen B cells (with I-A100% and I-A12% denoting B6 and I-A12%, respectively) were: I-A100%, 5295; I-A12%, 677; for spleen DCs: I-A100%, 5910; I-A12%, 652; for cTECs: I-A100%, 3544; I-A12%, 506; for mTECs: I-A100%, 3659, I-A12%, 516; and for thymic DCs: I-A100%, 4663, I-A12%, 517. (B) I-A100% and I-A12% spleen cells and thymocytes were stained as in (A); indicated subsets were obtained by cell sorting and then analyzed for I-Aα and I-Aβ mRNA expression by real-time PCR (forward [F] and reverse [R] primer positions as indicated). To avoid detection of I-A-1 and I-A-2, the forward primer spanned sequences from both exon 2 and exon 3. Data shown are representative of two independently performed experiments. *p ≤ 0.05, **p ≤ 0.01.

Close modal

Total thymocytes of I-A100% and I-A12% mice contained nearly identical percentages of CD4+CD8 T cells (Fig. 3A). In the spleen, the percentages of both CD4+ and CD8+ were higher for I-A12% than those of I-A100% mice (Fig. 3A), although CD4+/CD8+ T cell ratios were similar (1.92 for I-A100% and 1.83 for I-A12%). As B cell development can take place in the total absence of I-A (17), we were initially surprised to find an unexpected but highly significant reduction of B cells in I-A12% mice (p < 0.001) that contributed to the increase in the relative percentages of T cells. Similarly reduced numbers of B cells were previously reported for I-Aα–KO mice (18). The main difference between complete I-Aα–KO and the I-A12% mice we have generated is the total absence of CD4+ T cells in the I-Aα–KO mice and the normal numbers of CD4+ T cells in the I-A12% mice. Total cell numbers per thymus for each of the thymic DN, DP, CD4+CD8, and CD4CD8+ subsets were also highly similar between I-A100% and I-A12% mice (Fig. 3B), consistent with normal and unaltered intrathymic T cell development in I-A12% mice. As DP thymocytes that have engaged positive selection express CD69 (19), our finding of similar numbers of CD69+ DP thymocytes between I-A100% and I-A12% mice is consistent with a similar degree of positive selection. The highly similar nature of T cell development between I-A100% and I-A12% mice also extended to regulatory T (Treg) cells, as similar numbers of Foxp3+ Treg cells were found.

FIGURE 3.

Normal CD4+ T cell development in I-A12% mice. (A) Thymocytes and spleen cells from I-A100% and I-A12% mice were stained and analyzed as described in 2Materials and Methods. CD4+CD8 T cells, CD8+CD4 T cells, CD4+CD8+ thymocytes, and CD4CD8 thymocytes were gated as shown with the percentages of positive cells indicated. (B) Thymocytes and spleen cells from I-A100% and I-A12% mice of 1–3 mo age (n = 8, mean ± SD) were stained and analyzed according to 2Materials and Methods. Absolute cell numbers for indicated subsets (per entire thymus or entire spleen) are shown. The criteria used for the indicated subsets are: DN, CD48; DP, CD4+8+; CD69+DP, TCRβ+CD69+CD4+CD8+; CD4SP, CD4+CD8; CD8SP, CD4CD8+; Treg cells, CD25+Foxp3+CD4+8; B cells, IgM+. (C) Absolute numbers of AND (Vα11+Vβ3+CD4+CD8) and OT-II (Vα2+Vβ5+CD4+CD8) TCR Tg cells (per entire thymus or spleen) for I-A100% and I-A12% mice are shown (n = 3, mean ± SD). Cell numbers (per thymus) for AND Tg mice were: I-A100% background, 177.9 ± 11.6 × 106 total thymocytes, 102.5 ± 2.7 × 106 Tg TCR-α+β+ T cells; I-A12% background, 133.7 ± 7.6 × 106 total thymocytes, 55.0 ± 4.2 × 106 Tg TCR-α+β+ CD4+ T cells. Cell numbers (per thymus) for OT-II Tg mice were: I-A100% background, 233 ± 18.8 × 106 total thymocytes, 46.6 ± 3.0 × 106 Tg TCR-α+β+ T cells; I-A12% background, 211.8 ± 16.7 × 106 total thymocytes, 64.1 ± 2.7 × 106 Tg TCR-α+β+ T cells. (D) OT-II TCR thymocytes on I-A100% and I-A12% backgrounds were stained with BV510–anti-Vα2 (BD Biosciences), FITC–anti-Vβ5, PE–anti-CD4, A405–anti-CD8, and A647–anti-Vα8 or anti-Vα11. Vα8 and Vα11 expression on gated Vα2+Vβ5+CD4+CD8 cells are shown; the percentages of Vα8+ or Vα11+ cells (boundaries as marked) are: I-A100% background, 0.45% Vα8+, 0.48% Vα11+; I-A12% background, 0.62% Vα8+, 0.66% Vα11+. The y-axis range was adjusted to better reveal the relatively low frequencies of Vα8+ and Vα11+ cells. (E) Thymus and spleen cells from I-A100% and I-A12% mice were stained as described in 2Materials and Methods. After gating on CD4+CD8 T cells, the histogram displays of the marker expression as indicated are shown. (F) I-A100% and I-A12% thymocytes from neonate (7 d old) and adult (2 mo old) mice were stained as described in 2Materials and Methods. Percentages of different Vβ/Vα family usage by CD4+CD8 T cells are shown (n = 3, mean ± SD). Total cell numbers per thymus are: 7-d I-A100%, 75.6 ± 8.6 × 106; 7-d I-A12%, 72.7 ± 7.6 × 106; 2-mo I-A100%, 186.6 ± 11.6 × 106; 2-mo I-A12%, 182.7 ± 9.7 × 106. (G) I-A100% and I-A12% mice were given EdU continuously for 7 d (days 0 to 7, shaded in gray) as described in 2Materials and Methods. The percentages of EdU+ cells of the indicated subsets were determined on days 1, 7, 10, 14, 21, and 28 after implantation of EdU pump on day 0 (n = 2; mean ± SD). Data shown are representative of two independently performed experiments. *p ≤ 0.05, ***p ≤ 0.001.

FIGURE 3.

Normal CD4+ T cell development in I-A12% mice. (A) Thymocytes and spleen cells from I-A100% and I-A12% mice were stained and analyzed as described in 2Materials and Methods. CD4+CD8 T cells, CD8+CD4 T cells, CD4+CD8+ thymocytes, and CD4CD8 thymocytes were gated as shown with the percentages of positive cells indicated. (B) Thymocytes and spleen cells from I-A100% and I-A12% mice of 1–3 mo age (n = 8, mean ± SD) were stained and analyzed according to 2Materials and Methods. Absolute cell numbers for indicated subsets (per entire thymus or entire spleen) are shown. The criteria used for the indicated subsets are: DN, CD48; DP, CD4+8+; CD69+DP, TCRβ+CD69+CD4+CD8+; CD4SP, CD4+CD8; CD8SP, CD4CD8+; Treg cells, CD25+Foxp3+CD4+8; B cells, IgM+. (C) Absolute numbers of AND (Vα11+Vβ3+CD4+CD8) and OT-II (Vα2+Vβ5+CD4+CD8) TCR Tg cells (per entire thymus or spleen) for I-A100% and I-A12% mice are shown (n = 3, mean ± SD). Cell numbers (per thymus) for AND Tg mice were: I-A100% background, 177.9 ± 11.6 × 106 total thymocytes, 102.5 ± 2.7 × 106 Tg TCR-α+β+ T cells; I-A12% background, 133.7 ± 7.6 × 106 total thymocytes, 55.0 ± 4.2 × 106 Tg TCR-α+β+ CD4+ T cells. Cell numbers (per thymus) for OT-II Tg mice were: I-A100% background, 233 ± 18.8 × 106 total thymocytes, 46.6 ± 3.0 × 106 Tg TCR-α+β+ T cells; I-A12% background, 211.8 ± 16.7 × 106 total thymocytes, 64.1 ± 2.7 × 106 Tg TCR-α+β+ T cells. (D) OT-II TCR thymocytes on I-A100% and I-A12% backgrounds were stained with BV510–anti-Vα2 (BD Biosciences), FITC–anti-Vβ5, PE–anti-CD4, A405–anti-CD8, and A647–anti-Vα8 or anti-Vα11. Vα8 and Vα11 expression on gated Vα2+Vβ5+CD4+CD8 cells are shown; the percentages of Vα8+ or Vα11+ cells (boundaries as marked) are: I-A100% background, 0.45% Vα8+, 0.48% Vα11+; I-A12% background, 0.62% Vα8+, 0.66% Vα11+. The y-axis range was adjusted to better reveal the relatively low frequencies of Vα8+ and Vα11+ cells. (E) Thymus and spleen cells from I-A100% and I-A12% mice were stained as described in 2Materials and Methods. After gating on CD4+CD8 T cells, the histogram displays of the marker expression as indicated are shown. (F) I-A100% and I-A12% thymocytes from neonate (7 d old) and adult (2 mo old) mice were stained as described in 2Materials and Methods. Percentages of different Vβ/Vα family usage by CD4+CD8 T cells are shown (n = 3, mean ± SD). Total cell numbers per thymus are: 7-d I-A100%, 75.6 ± 8.6 × 106; 7-d I-A12%, 72.7 ± 7.6 × 106; 2-mo I-A100%, 186.6 ± 11.6 × 106; 2-mo I-A12%, 182.7 ± 9.7 × 106. (G) I-A100% and I-A12% mice were given EdU continuously for 7 d (days 0 to 7, shaded in gray) as described in 2Materials and Methods. The percentages of EdU+ cells of the indicated subsets were determined on days 1, 7, 10, 14, 21, and 28 after implantation of EdU pump on day 0 (n = 2; mean ± SD). Data shown are representative of two independently performed experiments. *p ≤ 0.05, ***p ≤ 0.001.

Close modal

The efficiency of T cell development in the context of T cells bearing monoclonal TCR was next examined. As shown in Fig. 3C, comparison of AND and OT-II TCR transgene-bearing CD4+ T cells on I-A100% and I-A12% genetic backgrounds revealed a 46% reduced development of AND Tg CD4+ T cells and a 38% increased development of OT-II Tg CD4+ T cells in I-A12% mice. OT-II Tg CD4+ T cells on both I-A100% and I-A12% backgrounds contained very few endogenous Vα8+ and Vα11+ T cells, consistent with positive selection of OT-II Tg TCR, rather than Tg TCRβ paired with endogenous Vα (Fig. 3D). Reduced I-A therefore exerted differential effects on the development of T cell–bearing different monoclonal TCRs.

Similar levels of TCRβ and CD5 expression levels were found for I-A100% and I-A12% CD4+ T cells (Fig. 3E), consistent with no alterations in the positive selection affinity. CD44 and Ly6C, often considered activation/memory marker, were also expressed at similar levels for I-A100% and I-A12% CD4+ T cells (Fig. 3E), again supporting the similar nature of CD4+ T cell development in I-A12% and I-A100% mice. CD69, Qa-2, and CD62L are markers that display differential expression between newly emerged CD4+CD8 thymocytes and their descendant peripheral CD4+ T cells. The differential expression patterns of CD69, Qa-2, and CD62L for thymus versus spleen CD4+ T cells for I-A12% mice were highly similar to those for I-A100% mice (Fig. 3E). Highly similar percentages of the 13 Vβ subfamilies and four Vα subfamilies as well as total number of CD4+ T cells were found for neonatal and adult I-A100% and I-A12% mice (Fig. 3F), clearly showing that CD4+ T cell development is normal-like in neonatal mice. The kinetics of CD4+ T cell development was monitored by a continuous 7-d EdU pulse, followed by chasing of EdU-labeled cells for an additional 21 d (Fig. 3G). After a 7-d continuous EdU labeling, nearly identical levels of EdU+ cells were found for DN and DP thymocytes of both I-A100% and I-A12% mice, indicating similar rates of entry into early thymocyte developmental stages for I-A100% and I-A12% mice. The expected slower kinetics of EdU incorporation into CD4+CD8 thymocytes was also similarly observed for I-A100% and I-A12% mice. Upon release of the 7-d EdU pulse, EdU+ DN and DP thymocytes rapidly turned over in both I-A100% and I-A12% mice, displaying no alteration in early T cell development kinetics for I-A12% mice. For both I-A100% and I-A12% CD4+CD8 thymocytes, there was a clear and similar level of increase of EdU+ cells 3 d after the release of EdU pulse (Fig. 3G, 10 d data), consistent with the rapid conversion of DP thymocytes into CD4+CD8 T cells. Thereafter, similar rates of turnover were seen for EdU+CD4+CD8 T cells in the thymus of I-A100% and I-A12% mice. Observation of EdU+ cells for a 21-d period after the release of the 7-d EdU pulse revealed that both I-A100% and I-A12% peripheral CD4+CD8 T cells displayed a similarly slow, steady turnover, indicating that the generally reduced I-A levels were not detrimental to the survival of CD4+ T cells in the periphery. Taken together, the development and survival of I-A12% CD4+ T cells were highly similar to normal (I-A100%) mice.

OT-II TCR (Vβ5+Vα2+) Tg mice express highly enriched I-A–restricted OVA-specific CD4+ T cells (20). We studied deletion of OT-II TCR-bearing CD4+ T cells in OT-II Tg mice on I-A100% and I-A12% backgrounds that either carried an OVA transgene or not (Fig. 4A). As expected, there were low percentages of Vβ5+Vα2+ cells in normal (nontransgenic) I-A100% and I-A12% mice that did not carry the OVA transgene. In mice harboring OT-II TCR but not the OVA transgene, highly elevated percentages of 85.9 and 90.1% Vβ5+Vα2+ cells within CD4+CD8 T cells were found for I-A100% and I-A12% mice, respectively, confirming the ability of ∼8-fold reduced I-A to positively select CD4+ T cells. In the presence of OVA, nearly all Vβ5+Vα2+ T cells disappeared in I-A100% and I-A12% mice, indicating high effectiveness of ∼8-fold reduced I-A expression in mediating intrathymic negative selection.

FIGURE 4.

Efficient negative selection in I-A12% mice demonstrated by a genetic approach. (A) Thymocytes taken from I-A100% and I-A12% mice carrying three different combinations of transgenes, that is, 1) OT-II TCR Tg alone, 2) OVA Tg alone, and 3) OT-II TCR Tg plus OVA Tg, were stained with allophycocyanin–Cy7–anti-CD4, A405–anti-CD8, FITC–anti-Vβ5, and A647-Vα2. CD4+CD8 and CD8+CD4 cells are marked as shown, with percentages of total cells falling within these gates shown. Correlated Vβ5/Vα2 contour plots of gated CD4+CD8 cells from indicated mice are shown on the lower panel, with percentages of Vβ5+Vα2+ cells shown. Data shown are representative of three independently performed experiments. (B) Panned spleen CD4+CD8 T cells from I-A100% and I-A12% mice were stimulated with sorted CD11c+ spleen cells obtained from C3H mice (left panel) or B6 mice (right panel), with or without added GK1.5 anti-CD4 mAb as described in 2Materials and Methods. Cell proliferation was measured on day 5 by a 6-h [3H]thymidine pulse. Data shown are representative of three independently performed experiments. (C) Lethally irradiated B6.TL mice (I-A100% H2bThy1a) were reconstituted with BM cells obtained from I-A0% (I-A–KO), I-A12%, and I-A100% mice, all Thy1b, as described in 2Materials and Methods. Donor-origin CD4+CD8Thy1.2+ T cells (2 × 105 cells per well) from chimeric mice were isolated by sorting, and their response to sorted allogeneic C3H CD11c+ DCs (1.2 × 105 cells per well, left panel) or to syngeneic B6 CD11c+ DC stimulator cells (1.2 × 105 cells per well, right panel) was studied. Cell proliferation was measured on day 5 by a 6-h [3H]thymidine pulse (n = 3, mean ± SD). Data shown are representative of two independently performed experiments. (D) MLR was set up using sorted C3H, I-A100%, and I-A12% spleen NK1TCRδCD4+8 T cells as responding T cells (2 × 105 cells per well) and indicated numbers of B6 (I-A100%) stimulator cells (T cell–depleted spleen cells). On day 5, cells from MLR cultures were recovered, stained with A405–anti-CD4, FITC–anti-CD25, and A647–anti-TCRβ, and analyzed by flow cytometry in the presence of CountBright absolute counting beads as described in 2Materials and Methods. Gates that mark counting beads and live cells (PI) are as indicated. Correlated CD4/CD25 contour plots of gated live CD4+TCRβ+ cells and percentages of CD4hiCD25+ cells (representing activated responding T cells) are shown. (E) Total number of activated CD4+ T cells per culture was calculated as described in 2Materials and Methods.

FIGURE 4.

Efficient negative selection in I-A12% mice demonstrated by a genetic approach. (A) Thymocytes taken from I-A100% and I-A12% mice carrying three different combinations of transgenes, that is, 1) OT-II TCR Tg alone, 2) OVA Tg alone, and 3) OT-II TCR Tg plus OVA Tg, were stained with allophycocyanin–Cy7–anti-CD4, A405–anti-CD8, FITC–anti-Vβ5, and A647-Vα2. CD4+CD8 and CD8+CD4 cells are marked as shown, with percentages of total cells falling within these gates shown. Correlated Vβ5/Vα2 contour plots of gated CD4+CD8 cells from indicated mice are shown on the lower panel, with percentages of Vβ5+Vα2+ cells shown. Data shown are representative of three independently performed experiments. (B) Panned spleen CD4+CD8 T cells from I-A100% and I-A12% mice were stimulated with sorted CD11c+ spleen cells obtained from C3H mice (left panel) or B6 mice (right panel), with or without added GK1.5 anti-CD4 mAb as described in 2Materials and Methods. Cell proliferation was measured on day 5 by a 6-h [3H]thymidine pulse. Data shown are representative of three independently performed experiments. (C) Lethally irradiated B6.TL mice (I-A100% H2bThy1a) were reconstituted with BM cells obtained from I-A0% (I-A–KO), I-A12%, and I-A100% mice, all Thy1b, as described in 2Materials and Methods. Donor-origin CD4+CD8Thy1.2+ T cells (2 × 105 cells per well) from chimeric mice were isolated by sorting, and their response to sorted allogeneic C3H CD11c+ DCs (1.2 × 105 cells per well, left panel) or to syngeneic B6 CD11c+ DC stimulator cells (1.2 × 105 cells per well, right panel) was studied. Cell proliferation was measured on day 5 by a 6-h [3H]thymidine pulse (n = 3, mean ± SD). Data shown are representative of two independently performed experiments. (D) MLR was set up using sorted C3H, I-A100%, and I-A12% spleen NK1TCRδCD4+8 T cells as responding T cells (2 × 105 cells per well) and indicated numbers of B6 (I-A100%) stimulator cells (T cell–depleted spleen cells). On day 5, cells from MLR cultures were recovered, stained with A405–anti-CD4, FITC–anti-CD25, and A647–anti-TCRβ, and analyzed by flow cytometry in the presence of CountBright absolute counting beads as described in 2Materials and Methods. Gates that mark counting beads and live cells (PI) are as indicated. Correlated CD4/CD25 contour plots of gated live CD4+TCRβ+ cells and percentages of CD4hiCD25+ cells (representing activated responding T cells) are shown. (E) Total number of activated CD4+ T cells per culture was calculated as described in 2Materials and Methods.

Close modal

As Tg OVA mice most likely express OVA at a level that is higher than endogenous peptides, we also studied intrathymic deletion in a more physiological context. Expression of I-A only on thymic cortex but not the medulla has been shown to result in the development of autoreactive T cells (21). CD4+ T cells that developed in I-A12% mice did not respond to syngeneic (I-Ab) stimulation, but responded vigorously to allogeneic (I-Ak) stimulation, indicating that autoreactive CD4+ T cells were deleted by intrathymic DCs and/or mTECs that express very low levels of I-A (Fig. 4B). Further study involved reconstitution of lethally irradiated I-A100% mice with I-A0%, I-A12%, and I-A100% bone marrow. As expected, CD4+ T cells from I-A0%→I-A100% chimeric mice proliferated in response to syngeneic I-A100% stimulation (Fig. 4C). Such syngeneic mixed lymphocyte response was not observed for I-A12%→I-A100% and I-A100%→I-A100% chimeric mice, again demonstrating deletion of autoreactive CD4+ T cells. A similarly potent allogeneic proliferative response against I-Ak was observed for CD4+ T cells that developed in I-A100% host mice reconstituted with I-A0%, I-A12%, and I-A100% bone marrow (Fig. 4C). Another way to examine whether reduced I-A expression inside the thymus can mediate deletion of CD4+ T cells capable of recognizing self in the context of I-A100% is to set up MLR cultures containing I-A12% CD4+ T cells and I-A100% APCs. To better allow visualization of low numbers of responding CD4+ T cells, we determined absolute numbers of activated CD4+ T cells based on increased CD4 and CD25 expression (22, 23). Live (PI) cells at the end of the 5-d MLR cultures were analyzed for correlated CD4/CD25 expression, with activated CD4+ T cells defined by elevated CD4 and CD25 expression (Fig. 4D). The presence of large numbers of activated CD25+CD4hi CD4+ T cells indicates potent activation of allogeneic C3H CD4+ T cells by I-Ab (I-A100%) stimulation. As expected, I-A100% CD4+ T cells responded poorly, if any, to I-A100% stimulator cells in that there were few live cells and among the live cells, as well as few CD25+CD4hi activated T cells (Fig. 4D). I-A12% responding CD4+ T cells also yielded few CD25+CD4hi T cells that were not significantly different from that of I-A100% responding CD4+ T cells, indicating that I-A12% mice do not contain more CD4+ T cells that are autoreactive against I-A100% APCs (Fig. 4D). Total absolute cell counts, a quantitative measure, of activated CD25+CD4hi T cells also showed comparably weak, if any, response by I-A12% and I-A100% CD4+ T cells against I-A100% APC stimulation, with allogeneic C3H CD4+ T cells showing potent activation (Fig. 4E).

Developing thymocytes that undergo negative selection express active caspase-3. We analyzed active caspase-3–expressing cells in CD4+CD8+ and CD4+CD8 thymocytes of I-A100% and I-A12% mice. CD69+CD4+CD8 thymocytes from I-A100% and I-A12% mice contained very similar numbers of caspase-3+ cells (Fig. 5A). CD69+CD4+CD8+ from I-A100% and I-A12% mice also contained very similar 0.131 and 0.125% caspase-3+ cells, respectively (Fig. 5A). Total counts of caspase-3+ cells per entire thymus were also highly similar when I-A12% and I-A100% mice were compared, regardless of whether CD69+CD4+CD8 or CD69+CD4+CD8+ subsets were analyzed (Fig. 5B). These results are consistent with no alteration in the degree of negative selection in I-A12% mice.

FIGURE 5.

Efficient negative selection in I-A12% mice revealed by biochemical and physical methods. (A) Thymocytes from indicated mice were stained as described in 2Materials and Methods. Gated CD4+CD8 and CD4+CD8+ subsets from I-A100% and I-A12% mice were examined for correlated CD69/caspase-3 expression. CD69+caspase-3+ cells are enclosed by a rectangular box with percentages of positive cells falling within the gate indicated. (B) Total counts of CD69+caspase-3+ cells within CD4+CD8 and CD4+CD8+ subsets per entire thymus were calculated and shown for I-A12% and I-A100% mice of 1 mo age (n = 4, mean ± SD). (C) Thymocytes from indicated mice were stained and subjected to multispectral imaging flow cytometric analysis as described in 2Materials and Methods. Through the use of the IDEAS analysis software, cells with fragmented nuclei were empirically determined to be most enriched in the “double low” region (shown as gated region) on a bivariate plot of aspect ratio intensity versus circularity morphology. Similar percentages of cells falling within this gated region (enriched for fragmented nuclei) were found for both of the CD4+CD8+ and CD4+CD8 subsets of I-A12% and I-A100% mice. (D) Representative bright field and Hoechst-stained nuclei images of normal cells and cells within the fragmented nuclei gate shown in (C) are displayed. (E) Total numbers of CD4+CD8+ and CD4+CD8 thymocytes (per entire thymus) with fragmented nuclei as defined in (C) were calculated and shown for I-A12% and I-A100% mice of 1 mo age (n = 2, mean ± SD). Data shown are representative of two independently performed experiments.

FIGURE 5.

Efficient negative selection in I-A12% mice revealed by biochemical and physical methods. (A) Thymocytes from indicated mice were stained as described in 2Materials and Methods. Gated CD4+CD8 and CD4+CD8+ subsets from I-A100% and I-A12% mice were examined for correlated CD69/caspase-3 expression. CD69+caspase-3+ cells are enclosed by a rectangular box with percentages of positive cells falling within the gate indicated. (B) Total counts of CD69+caspase-3+ cells within CD4+CD8 and CD4+CD8+ subsets per entire thymus were calculated and shown for I-A12% and I-A100% mice of 1 mo age (n = 4, mean ± SD). (C) Thymocytes from indicated mice were stained and subjected to multispectral imaging flow cytometric analysis as described in 2Materials and Methods. Through the use of the IDEAS analysis software, cells with fragmented nuclei were empirically determined to be most enriched in the “double low” region (shown as gated region) on a bivariate plot of aspect ratio intensity versus circularity morphology. Similar percentages of cells falling within this gated region (enriched for fragmented nuclei) were found for both of the CD4+CD8+ and CD4+CD8 subsets of I-A12% and I-A100% mice. (D) Representative bright field and Hoechst-stained nuclei images of normal cells and cells within the fragmented nuclei gate shown in (C) are displayed. (E) Total numbers of CD4+CD8+ and CD4+CD8 thymocytes (per entire thymus) with fragmented nuclei as defined in (C) were calculated and shown for I-A12% and I-A100% mice of 1 mo age (n = 2, mean ± SD). Data shown are representative of two independently performed experiments.

Close modal

Cells undergoing apoptosis display morphological changes such as nuclear condensation and fragmentation. Through multispectral imaging flow cytometry (24), we measured cells that displayed characteristic alterations in nuclear morphology (Fig. 5C). Through arbitrary bivariate analysis of the numerous data parameters, we identified that the vast majority of thymocytes that displayed nuclear shape alterations can best be visualized by bivariate analysis of circularity morphology versus aspect ratio intensity (Fig. 5C). Cells that are low in both circularity morphology and aspect ratio intensity (rectangular gated region, Fig. 5C) always contained cells with alterations in nuclear shape (Fig. 5D), whereas cells that fell within the bulk region away from the gate invariably displayed round nuclear morphology. The fraction of cells that fell within this gated region for I-A100% and I-A12% CD4+CD8 thymocytes were highly similar (Fig. 5C). For CD4+CD8+ thymocytes, the fraction of cells with altered nuclear morphology was also similar for I-A100% and I-A12% mice (Fig. 5C). Total counts of cells with altered nuclear morphology per entire thymus were again highly similar when I-A12% and I-A100% mice were compared, regardless of whether CD4+CD8 or CD4+CD8+ subsets were analyzed (Fig. 5E). These results are also consistent with little or no alteration in negative selection in I-A12% mice.

I-A100% and I-A12% spleen CD4+CD8 T cells were stimulated by anti-CD3, and nearly identical proliferative responses were observed across a wide range of anti-CD3 concentrations (Fig. 6A). IL-2 production in response to plate-bound anti-CD3/CD28 stimulation was also similar for I-A100% and I-A12% T cells (Fig. 6B). OT-II CD4+ T cells that developed in I-A100% and I-A12% mice were stimulated with graded doses of antigenic OVA peptide, and again nearly identical proliferation and IL-2 production responses were seen (Fig. 6C, 6D). DCs from I-A12% mice, owing to their highly reduced I-A expression, were highly deficient in stimulating allogeneic CD4+ T cell proliferation (Fig. 6E). The allogeneic response was strongly inhibited by anti-CD4 mAb, indicating the CD4 coreceptor–dependent nature. Upon addition of identical amounts of OVA, I-A12% spleen cells revealed a reduced ability to function as APCs in stimulating the proliferation of OT-II CD4+ T cells (Fig. 6F).

FIGURE 6.

Normal proliferation and IL-2 production by CD4+ T cells from I-A12% mice. (A) Triplicate cultures of panned spleen CD4+CD8 T cells (2 × 104 cells per well) from I-A100% and I-A12% mice were stimulated by indicated concentrations of anti-CD3 presented by B6 B cell blasts (105 cells per well). Cell proliferation was assessed on day 3 by a 6-h [3H]thymidine pulse. (B) Panned spleen CD4+CD8 T cells (105 cells per well) from I-A100% and I-A12% mice were stimulated by plate-bound anti-CD3/CD28 (10 μg/ml each), and IL-2 was released into day 3 culture supernatant determined by bioassay. Data shown are representative of two independently performed experiments. Sorted spleen Vα2+Vβ5+CD4+CD8 cells from OT-II TCR Tg mice on either I-A100% or I-A12% background were stimulated as described in 2Materials and Methods. Cell proliferation and IL-2 bioactivity released into day 3 culture supernatants are shown in (C) and (D), respectively. (E) Reduced allogeneic response of C3H CD4+ T cells to I-A12% DCs. Panned C3H spleen CD4+CD8 T cells were stimulated by sorted CD11c+ spleen cells from I-A100% or I-A12% mice as described in 2Materials and Methods. Cell proliferation was measured on day 5 by a 6-h [3H]thymidine pulse (n = 3; mean ± SD). (F) Sorted spleen Vα2+Vβ5+CD4+CD8 cells from OT-II TCR Tg mice on I-A100% background were set up in culture, to which indicated concentrations of OVA protein and B6 B cell blasts were added as described in 2Materials and Methods. Cell proliferation was assessed on day 3 by a 6-h [3H]thymidine pulse (n = 3; mean ± SD). Data shown are representative of two independently performed experiments. *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001.

FIGURE 6.

Normal proliferation and IL-2 production by CD4+ T cells from I-A12% mice. (A) Triplicate cultures of panned spleen CD4+CD8 T cells (2 × 104 cells per well) from I-A100% and I-A12% mice were stimulated by indicated concentrations of anti-CD3 presented by B6 B cell blasts (105 cells per well). Cell proliferation was assessed on day 3 by a 6-h [3H]thymidine pulse. (B) Panned spleen CD4+CD8 T cells (105 cells per well) from I-A100% and I-A12% mice were stimulated by plate-bound anti-CD3/CD28 (10 μg/ml each), and IL-2 was released into day 3 culture supernatant determined by bioassay. Data shown are representative of two independently performed experiments. Sorted spleen Vα2+Vβ5+CD4+CD8 cells from OT-II TCR Tg mice on either I-A100% or I-A12% background were stimulated as described in 2Materials and Methods. Cell proliferation and IL-2 bioactivity released into day 3 culture supernatants are shown in (C) and (D), respectively. (E) Reduced allogeneic response of C3H CD4+ T cells to I-A12% DCs. Panned C3H spleen CD4+CD8 T cells were stimulated by sorted CD11c+ spleen cells from I-A100% or I-A12% mice as described in 2Materials and Methods. Cell proliferation was measured on day 5 by a 6-h [3H]thymidine pulse (n = 3; mean ± SD). (F) Sorted spleen Vα2+Vβ5+CD4+CD8 cells from OT-II TCR Tg mice on I-A100% background were set up in culture, to which indicated concentrations of OVA protein and B6 B cell blasts were added as described in 2Materials and Methods. Cell proliferation was assessed on day 3 by a 6-h [3H]thymidine pulse (n = 3; mean ± SD). Data shown are representative of two independently performed experiments. *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001.

Close modal

Purified CD4+ T cells obtained from I-A100% and I-A12% mice were adoptively transferred into TCRβ-KO mice, followed by NP-CGG immunization with NP-CGG, a T-dependent Ag. Similar levels of anti-NP Ab responses were observed on day 16 and day 32 after immunization (Fig. 7A), indicating that I-A100% and I-A12% CD4+ T cells were comparable at providing help for Ab response.

FIGURE 7.

I-A12% CD4+ T cells possess normal Th and Treg cell function. (A) I-A100% or I-A12% CD4+ T cells were adoptively transferred into TCRβ-KO hosts, NP-CGG immunized 1 d later, and the levels of anti-NP IgG1 and IgG2a responses were determined on indicated days as described in 2Materials and Methods. NP-CGG immunized B6 and TCRβ-KO mice served as positive and negative controls, respectively (n = 5; mean ± SD). Preimmune serum contained negligible amounts of anti-NP IgG1 Abs (≤10 U/ml) and anti-NP IgG2a Abs (≤0.1 U/ml) for all mice. (B) Body weight of Rag1-KO hosts was determined at indicated times after receiving CD45RBhiCD4+CD25 (CD25) cells with or without CD45RBloCD4+CD25+ Treg cells as described in 2Materials and Methods, with body weight at time of cell transfer set as the 100% reference. (C) On day 60 after adoptive transfer or when body weight drops to <80% of the original weight, whichever is reached first, the mice were sacrificed and the colon (spanning the cecum to the anus) was removed and weighed (CD25 groups, n = 3; CD25 plus Treg cell groups, n = 4), with mean ± SD shown. (DJ) RAG1-KO hosts that had been adoptively transferred with I-A100% or I-A12% CD4+CD25CD45RBlo cells, as well as no cell transfer controls, were sacrificed at 7 wk after adoptive cell transfer. Serum levels of ALT (D) and AST (E) are shown. (G–J) Representative H&E images (original magnification, ×100) taken from indicated mice are shown, with representative lesions marked: open arrows indicate cell infiltration; asterisks denote necrotic hepatocytes. (F) Combined liver pathology scores based on the severity of cell infiltration and necrosis are shown. To facilitate correlation of ALT/AST levels with liver pathology, the mice marked nos. 1–4 denote the same individuals in (F)–(I). An F test for variance was analyzed for AST, ALT, and pathology score shown in subpanels (D)–(F). *p ≤ 0.05, **p ≤ 0.01.

FIGURE 7.

I-A12% CD4+ T cells possess normal Th and Treg cell function. (A) I-A100% or I-A12% CD4+ T cells were adoptively transferred into TCRβ-KO hosts, NP-CGG immunized 1 d later, and the levels of anti-NP IgG1 and IgG2a responses were determined on indicated days as described in 2Materials and Methods. NP-CGG immunized B6 and TCRβ-KO mice served as positive and negative controls, respectively (n = 5; mean ± SD). Preimmune serum contained negligible amounts of anti-NP IgG1 Abs (≤10 U/ml) and anti-NP IgG2a Abs (≤0.1 U/ml) for all mice. (B) Body weight of Rag1-KO hosts was determined at indicated times after receiving CD45RBhiCD4+CD25 (CD25) cells with or without CD45RBloCD4+CD25+ Treg cells as described in 2Materials and Methods, with body weight at time of cell transfer set as the 100% reference. (C) On day 60 after adoptive transfer or when body weight drops to <80% of the original weight, whichever is reached first, the mice were sacrificed and the colon (spanning the cecum to the anus) was removed and weighed (CD25 groups, n = 3; CD25 plus Treg cell groups, n = 4), with mean ± SD shown. (DJ) RAG1-KO hosts that had been adoptively transferred with I-A100% or I-A12% CD4+CD25CD45RBlo cells, as well as no cell transfer controls, were sacrificed at 7 wk after adoptive cell transfer. Serum levels of ALT (D) and AST (E) are shown. (G–J) Representative H&E images (original magnification, ×100) taken from indicated mice are shown, with representative lesions marked: open arrows indicate cell infiltration; asterisks denote necrotic hepatocytes. (F) Combined liver pathology scores based on the severity of cell infiltration and necrosis are shown. To facilitate correlation of ALT/AST levels with liver pathology, the mice marked nos. 1–4 denote the same individuals in (F)–(I). An F test for variance was analyzed for AST, ALT, and pathology score shown in subpanels (D)–(F). *p ≤ 0.05, **p ≤ 0.01.

Close modal

CD45RBloCD25CD4+ T cells from either I-A100% or I-A12% mice, when adoptively transferred into RAG-KO hosts, induced colitis development (Fig. 7B, 7C). CD45RBhiCD25+CD4+ Treg cells from either I-A12% or I-A100% mice prevented colitis development induced by CD45RBloCD25CD4+ T cells from I-A100% mice. CD45RBloCD25CD4+ T cells from I-A12% mice, however, displayed significantly more autoaggression against the liver than did their counterpart cells from I-A100% mice when adoptively transferred into RAG-KO hosts. All of the RAG-KO mice (n = 4; represented by mouse no. 4) that received adoptive transfer of CD45RBloCD25CD4+ T cells from I-A100% mice developed similar levels of liver damage as monitored by serum AST/ALT and by H&E-stained liver tissue sections (Fig. 7D–G). In two of five RAG-KO mice that received CD45RBloCD25CD4+ T cells from I-A12% mice, much more significant liver damage was observed. These two mice (marked as no. 1 and no. 2) had much more elevated serum AST/ALT (Fig. 7D, 7E), indicative of more severe damage, which was confirmed by pathology scoring based on infiltration and necrotic lesion (Fig. 7F–I). The degree of liver damage of the other three mice of this group (represented by mouse no. 3, Fig. 7D–F, 7H) was similar to RAG-KO mice that were adoptively transferred with I-A100% CD45RBloCD25CD4+ T cells in that they had lower serum AST/ALT; however, all had evidence of infiltration, but none showed necrotic lesions. RAG-KO control mice that did not receive any cell transfer developed neither colitis nor liver damage (Fig. 7D–F, 7J).

To examine the effect of reduced I-A expression on Ag-specific CD4+ T cell immune response, I-A12% and I-A100% mice were infected with sublethal doses of L. monocytogenes at 5000, 1000, and 200 CFU, followed by studying recall responses 5 wk later. When infected at a dose of 5000 L. monocytogenes CFU, I-A12% and I-A100% mice made similar recall IL-2 and IFN-γ production responses against the dominant LLO190 antigenic peptide (15) (Fig. 8A). For I-A100% mice, decreasing the dose of L. monocytogenes infection from 5000 to 1000 CFU had no effect on recall IL-2 production response and only a slight 23.4% reduction in IFN-γ production (Fig. 8A). The 5000 to 1000 CFU drop, however, had a much more striking effect on the I-A12% mice, with the recall IL-2 and IFN-γ production responses decreasing by 66.9 and 90.6%, respectively. At the lowest L. monocytogenes infection dose studied (200 CFU), readily detectable recall production of IL-2 and IFN-γ by I-A100% mice was still evident, but such production was highly deficient or absent for I-A12% mice.

FIGURE 8.

Weak Ag-specific CD4+ T cell recall response by I-A12% mice. I-A100% and I-A12% mice were infected (i.v.) by L. monocytogenes at indicated doses and sacrificed 5 wk postinfection. Sorted spleen CD4+ T cells were stimulated by (A) dominant antigenic peptide LLO190, (B) subdominant antigenic peptide LLO318 presented by B6 B cell blasts as described in 2Materials and Methods, and by (C) plate-bound 10 μg/ml each of anti-CD3 and anti-CD28 (105 cells per well). Cytokines released into day 2 culture supernatants were assayed by CBA (n = 2; mean ± SD). Data shown are representative of three independently performed experiments. For I-A100% and I-A12% mice infected with L. monocytogenes at a dose of 1000 CFU, the presence of GK1.5 anti-CD4 mAb (10 μg/ml) strongly inhibited recall cytokine responses against the LLO190 peptide (IL-2 by I-A100% mice, no GK1.5, 2560 pg/ml; with GK1.5, 30 pg/ml; IL-2 by I-A12% mice, no GK1.5, 180 pg/ml; with GK1.5, 28 pg/ml; IFN-γ by I-A100% mice, no GK1.5, 41.3 ng/ml; with GK1.5, 0.7 ng/ml; IFN-γ by I-A12% mice, no GK1.5, 6.6 ng/ml; with GK1.5, 0.2 pg/ml). (D) I-A100% and I-A12% mice were infected (i.v.) with L. monocytogenes (5000 CFU), sacrificed 5 wk postinfection, and their spleen cells were stained and analyzed by flow cytometry as described in 2Materials and Methods. Dot plots of correlated CD4/I-Ab–LLO190 tetramer expression and percentages of tetramer+ cells among gated CD4+CD8 T cells from one each of I-A100% and I-A12% mice that had previously been infected with 5000 CFU of L. monocytogenes and are shown. (E) Percentages of I-Ab–LLO190 tetramer+ cells among spleen CD4+CD8 T cells from I-A100% and I-A12% mice infected with 5000 CFU or 1000 CFU of L. monocytogenes are shown (n = 5; mean ± SD). (F) I-A100% and I-A12% mice were infected (intranasally) by influenza virus, sacrificed 5 wk postinfection, and their spleen cells were stained and analyzed by flow cytometry. Dot plots of correlated CD4/I-Ab–NP311 tetramer expression and percentages of tetramer+ cells within gated CD4+CD8 cells from one each of I-A100% and I-A12% mice that had previously been infected with influenza virus (104 PFU) are shown. (G) Percentages of I-Ab–NP311 tetramer+ cells within gated CD4+CD8 T cells from I-A100% and I-A12% mice infected with 104 PFU or 2 × 103 PFU of influenza virus are shown (n = 4; mean ± SD). Cells stained with control I-Ab/CLIP tetramer resulted in very few positive cells that ranged from 0.004 to 0.007%. *p ≤ 0.05, **p ≤ 0.01.

FIGURE 8.

Weak Ag-specific CD4+ T cell recall response by I-A12% mice. I-A100% and I-A12% mice were infected (i.v.) by L. monocytogenes at indicated doses and sacrificed 5 wk postinfection. Sorted spleen CD4+ T cells were stimulated by (A) dominant antigenic peptide LLO190, (B) subdominant antigenic peptide LLO318 presented by B6 B cell blasts as described in 2Materials and Methods, and by (C) plate-bound 10 μg/ml each of anti-CD3 and anti-CD28 (105 cells per well). Cytokines released into day 2 culture supernatants were assayed by CBA (n = 2; mean ± SD). Data shown are representative of three independently performed experiments. For I-A100% and I-A12% mice infected with L. monocytogenes at a dose of 1000 CFU, the presence of GK1.5 anti-CD4 mAb (10 μg/ml) strongly inhibited recall cytokine responses against the LLO190 peptide (IL-2 by I-A100% mice, no GK1.5, 2560 pg/ml; with GK1.5, 30 pg/ml; IL-2 by I-A12% mice, no GK1.5, 180 pg/ml; with GK1.5, 28 pg/ml; IFN-γ by I-A100% mice, no GK1.5, 41.3 ng/ml; with GK1.5, 0.7 ng/ml; IFN-γ by I-A12% mice, no GK1.5, 6.6 ng/ml; with GK1.5, 0.2 pg/ml). (D) I-A100% and I-A12% mice were infected (i.v.) with L. monocytogenes (5000 CFU), sacrificed 5 wk postinfection, and their spleen cells were stained and analyzed by flow cytometry as described in 2Materials and Methods. Dot plots of correlated CD4/I-Ab–LLO190 tetramer expression and percentages of tetramer+ cells among gated CD4+CD8 T cells from one each of I-A100% and I-A12% mice that had previously been infected with 5000 CFU of L. monocytogenes and are shown. (E) Percentages of I-Ab–LLO190 tetramer+ cells among spleen CD4+CD8 T cells from I-A100% and I-A12% mice infected with 5000 CFU or 1000 CFU of L. monocytogenes are shown (n = 5; mean ± SD). (F) I-A100% and I-A12% mice were infected (intranasally) by influenza virus, sacrificed 5 wk postinfection, and their spleen cells were stained and analyzed by flow cytometry. Dot plots of correlated CD4/I-Ab–NP311 tetramer expression and percentages of tetramer+ cells within gated CD4+CD8 cells from one each of I-A100% and I-A12% mice that had previously been infected with influenza virus (104 PFU) are shown. (G) Percentages of I-Ab–NP311 tetramer+ cells within gated CD4+CD8 T cells from I-A100% and I-A12% mice infected with 104 PFU or 2 × 103 PFU of influenza virus are shown (n = 4; mean ± SD). Cells stained with control I-Ab/CLIP tetramer resulted in very few positive cells that ranged from 0.004 to 0.007%. *p ≤ 0.05, **p ≤ 0.01.

Close modal

At the L. monocytogenes infection dose of 5000 CFU, recall IL-2 and IFN-γ production responses against the subdominant LLO318 antigenic peptide by I-A12% mice were 49% (38.7 ÷ 78.8 = 0.49) and 42% (0.63 ÷ 1.51 = 0.42) those of I-A100% mice, respectively (Fig. 8B). At an L. monocytogenes infection dose of 1000 CFU, recall IL-2 and IFN-γ production responses against the subdominant LLO318 antigenic peptide by I-A12% mice were 13.5% (7.5 ÷ 55.4 = 0.135) and 4.0% (0.038 ÷ 0.94 = 0.040) those of I-A100% mice, respectively. At the lowest L. monocytogenes infection dose of 200 CFU, recall IL-2 production response was reduced but nevertheless detectable at 31.1 ng/ml for I-A100% mice, whereas this response was extremely weak at 2.8 ng/ml for I-A12% mice. Recall IFN-γ production responses against the subdominant LL0318 peptide for both I-A100% and I-A12% infected with L. monocytogenes at a dose of 200 CFU were both negligible.

We next examined Listeria Ag-specific CD4+ T cells by tetramer staining of CD4+ T cells from I-A12% and I-A100% that had previously been infected with L. monocytogenes (Fig. 8D). Similar frequencies of I-Ab/LLO190–201-tetramer+CD4+ T cells were found for I-A12% and I-A100% mice that had been previously infected with a relatively high dose of L. monocytogenes (Fig. 8D), and for those that were infected at a lower L. monocytogenes infection dose, more tetramer+CD4+ T cells were found for I-A100% than for I-A12% mice (Fig. 8E). In addition to L. monocytogenes infection, we examined how I-A100% and I-A12% mice responded to influenza virus infection. A relatively high dose of influenza virus infection resulted in the development of similar numbers of Ag-specific CD4+ T cells as detected by I-Ab–NP311–325 tetramer staining (Fig. 8F, 8G). At a lower dose of influenza virus infection, significantly fewer Ag-specific CD4+ T cells were generated in I-A12% than in I-A100% mice.

Naive CD44loCD4+ and memory-type CD44hiCD4+ T cell subsets isolated from I-A12% and I-A100% mice (n = 4 per group) were examined for their ability to produce effector cytokines. Naive CD44loCD4+ T cells from both I-A12% and I-A100% mice produced IL-2 and low but detectable TNF-α and IL-6 in response to plate-bound anti-CD3/CD28 stimulation, but they did not produce significant levels of IFN-γ, IL-4, IL-5, IL-10, IL-13, and IL-17. IFN-γ and TNF-α are both considered Th1 type effector cytokines (25). Whereas the magnitude of IFN-γ production measured at both 24 and 48 h after anti-CD3/CD28 stimulation was similar for I-A12% and I-A100% CD44hiCD4+ memory cells, the TNF-α response was significantly higher for I-A12% than I-A100% memory CD4+ T cells (Fig. 9C). Of the Th2 cytokines examined, the amounts of IL-4, IL-5, IL-6, and IL-13 produced at 24 h for I-A12% memory CD4+ T cells were 5.0-, 1.7-, 2.0-, and 4.2-fold those of I-A100% memory CD4+ T cells, respectively. At 48 h, the amounts of IL-4, IL-5, IL-6, and IL-13 produced by I-A12% memory CD4+ T cells were 10.3-, 2.4-, 2.9-, and 4.2-fold those of I-A100% memory CD4+ T cells, respectively. After subjecting these data to a Student t test, significant statistical differences for IL-4 production at 24 h and for IL-6 and IL-13 production at both 24 and 48 h were found (Fig. 9D, 9F, 9H). IL-10, also considered a Th2 cytokine, was different in that its production displayed little difference between I-A12% and I-A100% memory CD4+ T cells (Fig. 9G). IL-17 production responses at 24 and 48 h for I-A12% memory CD4+ T cells were 3.8- and 3.6-fold those of I-A100% memory CD4+ T cells, respectively, with the 48-h response showing a significant statistical difference (Fig. 9I).

FIGURE 9.

Increased inducibility of many cytokine genes in I-A12% memory-like CD44hiCD4+ T cells. Sorted I-A12% and I-A100% spleen naive CD4+ T cells as defined by CD44loNK1TCRδCD4+CD8 marker expression and memory CD4+ T cells as defined by CD44hiNK1TCRδCD4+CD8 marker expression were set up in triplicate cultures that were precoated with anti-CD3/CD28 (with uncoated wells as controls). NKT cells and TCR-γδ T cells, both of which can produce significant amounts of many cytokines, were removed by cell sorting. Cytokines secreted by 1 d– and 2 d–stimulated cells were assayed by CBA: (A) IL-2, (B) IFN-γ, (C) TNF-α, (D) IL-4, (E) IL-5, (F) IL-6, (G) IL-10, (H) IL-13, and (I) IL-17 (n = 4; mean ± SD). (J) Sorted spleen CD44hiCD4+CD8 T cells (105 cells per well) from mice infected with 5000 CFU of L. monocytogenes 5 wk previously were stimulated by LLO190 peptide presented by B6 B cell blasts (2 × 105 cells per well). Indicated cytokines released into day 2 culture supernatants were assayed by CBA (n = 4; mean ± SD). *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001. ND, not detected.

FIGURE 9.

Increased inducibility of many cytokine genes in I-A12% memory-like CD44hiCD4+ T cells. Sorted I-A12% and I-A100% spleen naive CD4+ T cells as defined by CD44loNK1TCRδCD4+CD8 marker expression and memory CD4+ T cells as defined by CD44hiNK1TCRδCD4+CD8 marker expression were set up in triplicate cultures that were precoated with anti-CD3/CD28 (with uncoated wells as controls). NKT cells and TCR-γδ T cells, both of which can produce significant amounts of many cytokines, were removed by cell sorting. Cytokines secreted by 1 d– and 2 d–stimulated cells were assayed by CBA: (A) IL-2, (B) IFN-γ, (C) TNF-α, (D) IL-4, (E) IL-5, (F) IL-6, (G) IL-10, (H) IL-13, and (I) IL-17 (n = 4; mean ± SD). (J) Sorted spleen CD44hiCD4+CD8 T cells (105 cells per well) from mice infected with 5000 CFU of L. monocytogenes 5 wk previously were stimulated by LLO190 peptide presented by B6 B cell blasts (2 × 105 cells per well). Indicated cytokines released into day 2 culture supernatants were assayed by CBA (n = 4; mean ± SD). *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001. ND, not detected.

Close modal

We next examined the ability of CD44hiCD4+ T cells from I-A100% and I-A12% mice 5 wk after L. monocytogenes infection to mount Ag-specific recall Th1, Th2, and Th17 cytokine production responses. Strong IFN-γ and TNF-α recall responses of similar magnitude were observed for I-A100% and I-A12% mice, consistent with L. monocytogenes being a strong Th1 inducer (Fig. 9J). Neither IL-4 nor IL-17 was detected for L. monocytogenes–infected I-A100% and I-A12% mice. The lack of production of the hallmark Th2 cytokine IL-4 was accompanied by clearly detectable but relatively low level production of other Th2 cytokines (IL-5, IL-6, IL-10, and IL-13) at levels that were similar for I-A100% and I-A12% mice.

In the absence of MHC-II, the development of conventional CD4+ Th cells is completely halted (1, 2). MHC-II therefore plays an indispensable role in CD4+ T cell development. However, just how much MHC-II is required for optimal CD4+ T cell development and immune response is unclear. Conserved high-level surface MHC-II expression by cTECs, mTECs, and DCs implies evolutionary survival advantages, and lowering of MHC-II expression is expected to be detrimental. To address this possibility, we established an ENU mutant mouse, named I-A12%, with a generalized ∼8-fold decreased MHC-II expression in cTECs, mTECs, and DCs. The I-A12% mouse carries an ENU-induced point mutation located in the branch site of intron 2 of the I-Aα gene. This point mutation replaces the conserved A of the branch site with a T, thus causing a splicing error in most of the I-Aα pre-mRNA with a consequential ∼8-fold reduced expression of wild-type mature I-Aα mRNA as well as cell surface I-Aα in cTECs, mTECs, DCs, and B cells. To our surprise, the quantity of Th and Treg cells that developed intrathymically was not reduced in I-A12% mice. Equally surprising was the normal-like functional capabilities of I-A12% Th and Treg cells. In the context of intrathymic T cell development, we have so far not clearly ascertained the reason behind the survival advantage presumably associated with high-level I-A expression by thymic cTECs, mTECs, and DCs. Our finding of increased autoaggression against the liver by CD4+ T cells of I-A12% origin is suggestive of a role of high I-A expression in mediating more complete deletion of certain tissue-specific autoreactive CD4+ T cells. Additionally, we have clearly found two deficits or alterations in peripheral CD4+ T cell immune response. First, the magnitude of the CD4+ immune response is significantly decreased in I-A12% mice, most likely the result of the reduced presentation of processed antigenic peptides due to reduced I-A expression on DCs. Second, memory-like CD4+ T cells in I-A12% mice showed elevated inducibility in TNF-α, IL-4, IL-6, IL-13, and IL-17 genes, but not in IL-2, IFN-γ, and IL-10 genes. TNF-α and IFN-γ are both considered Th1 cytokines; however, TNF-α but not IFN-γ showed upregulation. All Th2 cytokine genes studied displayed increased inducibility, except for IL-10. Thus, the increased inducibility of cytokine genes is not correlated with the Th1 or Th2 pattern. What is clear is that the relative ratios of the various cytokines that are produced upon TCR stimulation of I-A12% memory CD4+ T cells can be quite distinct from those of I-A100% mice. On a gross level, I-A12% mice are normal and reproduce well, and histological staining of major tissues and organs has not revealed any evidence of cellular infiltration. There is not an apparent hole in the repertoire of CD4+ T cells that developed under conditions of low I-A, as similar numbers of Ag-specific CD4+ T cells were found for I-A12% and I-A100% mice after they were infected with adequate doses of L. monocytogenes or influenza virus. Additionally, I-A12% and I-A100% CD4+ T cells were equally effective in providing help to B cells in a T-dependent NP-CGG Ab response. Concerning possible alterations in the TCR repertoire, whereas I-A12% and I-A100% CD4+CD25 T cells were comparable in inducing colitis when transferred into a RAG-KO host, I-A12% CD4+CD25 T cells displayed stronger aggression against the liver. Because liver-specific autoreactive CD4+ T cells undergo intrathymic deletion due to Aire-dependent expression of tissue-specific Ags (26), this finding is suggestive of lower expression levels of liver tissue-specific Ags than gut-specific Ags by Aire+ mTECs, thus resulting in impaired deletion of liver-autoreactive CD4+ T cells compared with gut-autoreactive CD4+ T cells. As for MLRs against B6 (I-A100%) APCs, both I-A12% and I-A100% CD4+ T cells responded poorly and comparably. DCs are the most potent stimulators of MLRs. Because the level of autoantigens expressed by DCs is much higher in quantity than that of tissue-specific Ags by mTECs (26), an ∼8-fold drop in I-A expression in DCs may still allow sufficient expression of DC-encoded self-peptides that results in effective deletion of DC-reactive CD4+ T cells. Owing to the low level Aire-driven tissue-specific Ag expression in mTECs, an ∼8-fold drop in I-A is likely too severe to efficiently mediate deletion, thus resulting in increased escape of liver-specific CD4+ T cells into the periphery.

Whether the two alterations in I-A12% mice noted above will result in significant detrimental outcome when placed in the context of known autoimmune mouse models or subjected to challenge by autoimmune response induction protocols requires further investigation.

Lowering the amount of I-A expression on cTECs, without affecting the peptide generation machinery, might be expected to result in the reduction of the number of peptides that can be effectively presented by cTECs to mediate positive selection. Results we report in the present study are surprising in that no deficiency in cellularity, function, and marker expression of CD4+ T cells in I-A12% mice were found, although a highly reproducible reduction in B cell numbers was observed. The reason behind reduced B cells is not fully understood, particularly because B cells develop normally in mice completely devoid of I-Aα and I-Aβ genes (17). However, reduced B cell number has previously been reported for I-Aα–KO mice, and unfolded protein response due to an excess of unfolded I-Aβ has been suggested as the likely cause (18). A similarly reduced B cell phenotype was reported in invariant chain–deficient mice, and this altered phenotype was corrected by I-Aβ deletion, providing strong support for the role of unpaired I-Aβ as the culprit in inducing unfolded protein response that then leads to impaired B cell survival (27).

Lineage-specific knockdown of I-A expression by ∼10-fold in mTECs but not DCs has been reported to cause an increase in CD4+CD8 T cells attributed to deficient negative selection (28). Reduced I-A expression on both mTECs and DCs, both of which play critical roles in negative selection, is characteristic of the I-A12% mice we have established. Because we found no evidence of deficient negative selection by three independent assays, we conclude that the highly reduced I-A expression in I-A12% mice is fully capable of mediating highly efficient negative selection. To reconcile the reduced negative selection when I-Aα is decreased by ∼10-fold in mTECs and our results showing no effect on negative selection when I-Aα is decreased by ∼8-fold in both mTECs and DCs, it is possible that an ∼8-fold decreased I-A expression is just sufficient enough for negative selection and that a slight further decrease would cause significant defects in negative selection. K14 promoter-driven expression of I-Aβb in I-Aβb-null mice allows the development of polyclonal autoreactive CD4+ T cells, and such autoreactive CD4+ T cells are deleted by lysozyme promoter-driven low level I-Aβb in macrophages and thymic DCs (29). Such lysozyme-driven low-level I-Aβb expression was not found in UEA+ thymic cells, indicating little or no expression on UEA+ mTECs. The I-A12% mice we have established differ in that they display reduced I-A expression on both DCs and mTECs. Regardless, our finding of highly efficient negative selection in I-A12% mice is consistent with restoration of negative selection by low I-A expression on thymic DCs (29).

The relative level of CD5 expression on CD4+ T cells reflects the affinity of positive selection, with low and high affinity resulting in CD5lo and CD5hi CD4+ T cells, respectively (30). Because CD4+ T cells that develop in I-A12% and I-A100% mice express similar levels of CD5, it would appear that there was no alteration of relevant positive selection signal forces. In this context, other markers expressed by CD4+ T cells such as CD62L, CD44, and Qa-2 were all nearly identical, indicating the highly similar nature of CD4+ T cell development in I-A12% and I-A100% mice.

CD4+ T cells generated from I-A12% mice effectively helped B cells to produce Abs against NP-CGG immunization. Treg cells that developed in I-A12% mice were also fully functional in preventing colitis initiated by CD4+CD25 T cells that had been adoptively transferred into RAG-KO hosts. The development of Treg cells has been reported to require stronger interaction between TCR and pMHC (31). An ∼8-fold decrease in I-A expression is expected to result in quantitatively less peptide presented to developing Treg cells, which prompted us initially to anticipate a decrease in the number of Treg cells that would develop in I-A12% mice. However, this expected drop in Treg cells was not found. It is possible that the strong signaling received by developing Treg cells results from a combination of TCR signaling and signaling mediated by other receptor–ligand interactions and that these other receptor–ligand interactions are the major contributors to the strong signaling that has been reported for Treg cells (31). The highly diverse and plastic nature of TCRs generated by random combinatorial joining of V-(D)-J genes also can provide a plausible explanation for the unaffected Treg cell generation in I-A12% mice, although Treg cells generated under conditions of limiting I-A might be expected to express an altered TCR repertoire in the context of Ag recognition that might affect the ability of I-A12% Treg cells to prevent colitis development induced by I-A100% CD4+CD25 T cells. Contrary to this expectation, we found that I-A12% Treg cells were fully capable of blocking colitis development induced by either I-A12% or I-A100% T cells. Further work is required to understand the apparent lack of adverse effect of reduced I-A expression on CD4+ T cell development and function. It can however be concluded that the reason for the conserved high MHC-II expression on cTECs, mTECs, and DCs is not related to the quality and quantity of CD4+ T cells using the common assays we have employed, although it is possible that I-A12% CD4+ T cells differ from I-A100% CD4+ T cells in ways that have escaped our observation.

Both autophagy-dependent and -independent peptides are loaded onto I-A in TECs (32). When AND TCR Tg mice were bred onto an I-A12% background, we observed a 46% reduction in the total number of AND TCR Tg CD4+ T cells that developed. Because the development of AND TCR Tg CD4+ T cells was autophagy-independent (32), but was adversely affected by low I-A, it is likely that the reduced I-A on cTECs resulted in preferential presentation of autophagy-dependent peptides. Further study of CD4+ T cell development in I-A12% mice in the context of autophagy-deficient background might reveal valuable insight.

Memory phenotype CD4+ T cells found in mice that had not been subjected to intentional immunization presumably develop from immune response to environmental Ags. This notion is supported by the finding that the vast majority of CD4+ T cells of germ-free mice are of the CD45Rhi (naive) type (33, 34), and significant numbers of CD45RloCD4+ T cells (Ag-experienced) arise after exposure to conventional (non–germ-free) housing conditions (33). Spontaneously arisen memory-like CD4+ T cells, but not naive CD4+ T cells, rapidly produce effector cytokines such as IL-4, IL-5, IL-6, and IFN-γ when stimulated (35, 36). The results we report in the present study confirm IL-4, IL-5, IL-6, and IFN-γ inducibility in memory-like but not naive CD4+ T cells from B6 (I-A100%) mice, and they further extend similarly increased inducibility to TNF-α, IL-10, IL-13, and IL-17 cytokine genes. Thus, when compared with naive CD4+ T cells, eight cytokine genes showed increased inducibility in B6 (I-A100%) memory-like CD4+ T cells. For I-A12% CD4+ T cells, five (TNF-α, IL-4, IL-6, IL-13, and IL-17) of these eight cytokine genes displayed further increased inducibility, with IFN-γ and IL-10 remaining at similar inducibility levels. The mechanism responsible for the heightened inducibility of TNF-α, IL-4, IL-6, IL-13, and IL-17 genes, but not IL-2, IFN-γ, and IL-10 genes, in I-A12% CD4+ T cells is unclear. Overall, the heightened cytokine inducibility profile of I-A12% memory CD4+ T cells is not fully concordant with the common Th1 and Th2 classification schema. Both TNF-α and IFN-γ are considered Th1 cytokines; however, I-A12% memory CD4+ T cells showed elevated TNF-α but not IFN-γ production, which might be caused by the expression of TNF-α–specific enhancing mechanisms in I-A12% but not I-A100% memory-like CD4+ T cells. Memory CD4+ T cells from I-A12% mice displayed increased inducibility in all of the Th2 cytokines examined, with the exception of IL-10. As low antigenic exposure has previously been shown to favor Th2 development in the context of monoclonal TCR Tg T cells (37), the increased Th2 cytokine inducibility in I-A12% memory CD4+ T cells may be caused by a lower degree of antigenic exposure. The reason for the discordant IL-10 inducibility is unclear but may be due to IL-10–specific regulatory mechanisms that do not negatively impact other Th2 cytokine genes. In vivo–generated memory CD4+ T cells may or may not follow Th1 and Th2 cytokine inducibility patterns, which were originally established for in vitro–passaged Th cell lines and clones (25, 38). It is also possible that cytokine inducibility patterns of in vivo–generated memory CD4+ T cells are more complex and that additional levels of regulation exist for the control of distinct Th1 and Th2 cytokine genes than is currently understood. The elevated inducibility of IL-4 and IL-13 in polyclonal memory CD4+ T cells may be relevant to the induction of allergies, as allergens are part of the environmental Ags to which humans are exposed.

Ag (L. monocytogenes)–specific CD4+CD44hi T cell recall response as measured by the production of Th1, Th2, and Th17 cytokines was highly similar for L. monocytogenes–immunized I-A12% and I-A100% mice. This result is in marked contrast to the highly elevated inducibility of many cytokines by endogenous memory-like CD4+CD44hi T cells of I-A12% mice. It is well known that L. monocytogenes infection results in a potent Th1 effector response. Endogenous CD4+CD44hi T cells, alternatively, may be a composite of Ag-specific Th1, Th2, and Th17 cells. Further study of Ags known to induce Th2 and Th17 recall responses may reveal new insight. Alternatively, the cytokine inducibility profile may change as a function of the age of CD4+ memory T cells (39). Another possibility is that a significant subset of endogenous CD4+CD44hi T cells is generated through the process of homeostatic expansion (40) and may therefore display significantly different functional capabilities. Further investigations are required to clarify the relative contribution of each of these possibilities to the highly elevated production of multiple cytokines induced by polyclonal activation of endogenous CD4+CD44hi T cells from I-A12% mice.

As reduced I-A expression is expected to result in quantitatively less presentation of antigenic peptides derived from Listeria, a subdued recall response is expected. This expectation is consistent with our finding of reduced CD4+ T cell recall response by I-A12% mice previously subjected to limiting L. monocytogenes infection. At limiting doses of L. monocytogenes infection, Ag-specific CD4+ T cells are expected to encounter lower doses of antigenic peptide during their initial activation, clonal expansion, and transition into memory cells, in the context of I-Ab in I-A12% mice. Because memory CD4+ T cells exposed to low doses of agonist peptides in the context of MHC-II have been shown to transition into a state of dormancy in which their effector function is significantly dampened (39, 4143), it is possible that the reduced CD4+ T cell response we have observed in I-A12% mice infected with relative low doses of L. monocytogenes is the result of a low-level agonist-induced state of dormancy. This possibility appears to be at odds with our finding that the reduced functional recall response was accompanied by reduced tetramer+CD4+ T cells. Alternatively, low level agonist-induced dormant CD4+ T cells are characterized by extended longevity, and after an extended period (>1 y) of dormancy, they respond to Ag re-encounter by potent cytokine release (39). The highly elevated cytokine inducibility of memory-like CD4+CD44hi T cells from I-A12% mice that had not been intentionally immunized may be due to dormant CD4+ T cells that were generated by an immune response to relatively low doses of environmental Ags. Further experiments designed to specifically address the potential induction of dormant memory CD4+ T cells in I-A12% mice are warranted.

The reduced I-A expression in I-A12% mice may influence memory cell generation as a result of Ag-stimulated CD4+ T cell activation or lymphopenia-induced homeostatic expansion, as both of these processes are dependent on I-A (44). Whether the highly reduced level of I-A expression in I-A12% mice also affects CD4+ T cell homeostatic expansion is unknown. We do know that memory-like CD4+ T cells constitute a similar fraction of CD4+ T cells in both I-A100% and I-A12% mice, although such a result does not directly address the mechanism responsible for their generation. CD8+ T cells that have undergone homeostatic expansion produce significantly less IFN-γ than do those that were true memory cells induced by Ag stimulation (45). If this finding can be applied to CD4+ T cells, then our finding of similar IFN-γ production for I-A100% and I-A12% memory-like CD4+ T cells would be consistent with a similar pathway (Ag or lymphopenia driven) that led to their generation. Increased IL-6 but reduced IL-4 production by spleen cells has been reported for CD4+ T cells that have undergone homeostatic expansion in lymphopenic hosts (46). Because we found increased inducibility in both IL-4 and IL-6 genes by I-A12% memory-like CD4+ T cells, it is likely that the memory-like CD4+ T cells in I-A12% mice were generated in response to stimulation by environmental Ags rather than by homeostatic expansion.

Finally, and in a more general sense, I-A–expressing cTECs, mTECs, and DCs are highly relevant in CD4+ T cell development, function, and response. As the I-A12% mouse displays reduced I-A expression on cTECs, mTECs, and DCs, its general availability will enable investigation on how limiting I-A affects normal as well as pathogenic CD4+ T cell immune responses.

We thank Gary Wu for ENU injection and Ya-Min Lin for performing excellent cell sorting. We also thank Ming-Chu Wang and the Institute of Molecular Biology Animal Facility staff for excellent animal husbandry.

This work was supported by National Science Council Grants NSC96-3112-B-001-007 and NSC96-3112-B-001-008, Ministry of Science and Technology Grant 105-2320-B-001-007, and by Academia Sinica Grants AS-91-IMB1PP and AS-95-TP-B06.

Abbreviations used in this article:

A

Alexa Fluor

ALT

alanine aminotransferase

AST

aspartate aminotransferase

B6

C57BL/6

BV

Brilliant Violet

CBA

cytokine bead array

cTEC

cortical thymic epithelial cell

DC

dentritic cell

DN

double-negative

DP

double-positive

EdU

5′-ethynyl-2′-deoxyuridine

FSC

forward light scatter

I-A

I-Aα:I-Aβ

I-E

I-Eα:I-Eβ

KO

knockout

LLO

listeriolysin O

MFI

mean fluorescence intensity

MHC-II

MHC class II

mTEC

medullary TEC

NP

4-hydroxy-3-nitrophenylacetyl

NP-CGG

4-hydroxy-3-nitrophenylacetyl hapten–conjugated chicken γ-globulin

PI

propidium iodide

SP

single-positive

SSC

side light scatter

TEC

thymic epithelial cell

Tg

transgenic

Treg

regulatory T.

1
Grusby
M. J.
,
Johnson
R. S.
,
Papaioannou
V. E.
,
Glimcher
L. H.
.
1991
.
Depletion of CD4+ T cells in major histocompatibility complex class II-deficient mice.
Science
253
:
1417
1420
.
2
Cosgrove
D.
,
Gray
D.
,
Dierich
A.
,
Kaufman
J.
,
Lemeur
M.
,
Benoist
C.
,
Mathis
D.
.
1991
.
Mice lacking MHC class II molecules.
Cell
66
:
1051
1066
.
3
Klein
L.
,
Kyewski
B.
,
Allen
P. M.
,
Hogquist
K. A.
.
2014
.
Positive and negative selection of the T cell repertoire: what thymocytes see (and don’t see).
Nat. Rev. Immunol.
14
:
377
391
.
4
Otten
L. A.
,
Steimle
V.
,
Bontron
S.
,
Mach
B.
.
1998
.
Quantitative control of MHC class II expression by the transactivator CIITA.
Eur. J. Immunol.
28
:
473
478
.
5
Reith
W.
,
LeibundGut-Landmann
S.
,
Waldburger
J. M.
.
2005
.
Regulation of MHC class II gene expression by the class II transactivator.
Nat. Rev. Immunol.
5
:
793
806
.
6
Mombaerts
P.
,
Clarke
A. R.
,
Rudnicki
M. A.
,
Iacomini
J.
,
Itohara
S.
,
Lafaille
J. J.
,
Wang
L.
,
Ichikawa
Y.
,
Jaenisch
R.
,
Hooper
M. L.
, et al
.
1992
.
Mutations in T-cell antigen receptor genes α and β block thymocyte development at different stages.
Nature
360
:
225
231
.
7
Kaye
J.
,
Hsu
M. L.
,
Sauron
M. E.
,
Jameson
S. C.
,
Gascoigne
N. R.
,
Hedrick
S. M.
.
1989
.
Selective development of CD4+ T cells in transgenic mice expressing a class II MHC-restricted antigen receptor.
Nature
341
:
746
749
.
8
Chang
J. F.
,
Thomas
C. A.
 III
,
Kung
J. T.
.
1991
.
Induction of high level IL-2 production in CD4+8 T helper lymphocytes requires post-thymic development.
J. Immunol.
147
:
851
859
.
9
Justice
M. J.
,
Noveroske
J. K.
,
Weber
J. S.
,
Zheng
B.
,
Bradley
A.
.
1999
.
Mouse ENU mutagenesis.
Hum. Mol. Genet.
8
:
1955
1963
.
10
Neuhaus
I. M.
,
Beier
D. R.
.
1998
.
Efficient localization of mutations by interval haplotype analysis.
Mamm. Genome
9
:
150
154
.
11
Gray
D. H.
,
Chidgey
A. P.
,
Boyd
R. L.
.
2002
.
Analysis of thymic stromal cell populations using flow cytometry.
J. Immunol. Methods
260
:
15
28
.
12
Chen
Y. T.
,
Kung
J. T.
.
2005
.
CD1d-independent developmental acquisition of prompt IL-4 gene inducibility in thymus CD161(NK1)CD44lowCD4+CD8 T cells is associated with complementarity determining region 3-diverse and biased Vβ2/Vβ7/Vβ8/Vα3.2 T cell receptor usage.
J. Immunol.
175
:
6537
6550
.
13
Chen
F. L.
,
Kung
J. T.
.
1996
.
Deficient CD4+ T cell proliferation in the class 1 MHC-restricted 2C TCR-transgenic mouse.
J. Immunol.
156
:
2036
2044
.
14
Lee
C. C.
,
Kung
J. T.
.
2012
.
Marginal zone B cell is a major source of Il-10 in Listeria monocytogenes susceptibility.
J. Immunol.
189
:
3319
3327
.
15
Geginat
G.
,
Schenk
S.
,
Skoberne
M.
,
Goebel
W.
,
Hof
H.
.
2001
.
A novel approach of direct ex vivo epitope mapping identifies dominant and subdominant CD4 and CD8 T cell epitopes from Listeria monocytogenes.
J. Immunol.
166
:
1877
1884
.
16
Hoffmann
E.
,
Neumann
G.
,
Kawaoka
Y.
,
Hobom
G.
,
Webster
R. G.
.
2000
.
A DNA transfection system for generation of influenza A virus from eight plasmids.
Proc. Natl. Acad. Sci. USA
97
:
6108
6113
.
17
Madsen
L.
,
Labrecque
N.
,
Engberg
J.
,
Dierich
A.
,
Svejgaard
A.
,
Benoist
C.
,
Mathis
D.
,
Fugger
L.
.
1999
.
Mice lacking all conventional MHC class II genes.
Proc. Natl. Acad. Sci. USA
96
:
10338
10343
.
18
Rolink
A. G.
,
Brocker
T.
,
Bluethmann
H.
,
Kosco-Vilbois
M. H.
,
Andersson
J.
,
Melchers
F.
.
1999
.
Mutations affecting either generation or survival of cells influence the pool size of mature B cells.
Immunity
10
:
619
628
.
19
Brändle
D.
,
Müller
S.
,
Müller
C.
,
Hengartner
H.
,
Pircher
H.
.
1994
.
Regulation of RAG-1 and CD69 expression in the thymus during positive and negative selection.
Eur. J. Immunol.
24
:
145
151
.
20
Barnden
M. J.
,
Allison
J.
,
Heath
W. R.
,
Carbone
F. R.
.
1998
.
Defective TCR expression in transgenic mice constructed using cDNA-based α- and β-chain genes under the control of heterologous regulatory elements.
Immunol. Cell Biol.
76
:
34
40
.
21
Laufer
T. M.
,
DeKoning
J.
,
Markowitz
J. S.
,
Lo
D.
,
Glimcher
L. H.
.
1996
.
Unopposed positive selection and autoreactivity in mice expressing class II MHC only on thymic cortex.
Nature
383
:
81
85
.
22
Dobrzanski
M. J.
,
Yang
T. J.
.
1992
.
Elevated CD4 antigen expression among activated T cells in lymph nodes draining mammary regions chronically exposed to Staphylococcus aureus antigen.
Immunobiology
185
:
53
62
.
23
Waldmann
T. A.
1991
.
The interleukin-2 receptor.
J. Biol. Chem.
266
:
2681
2684
.
24
Henery
S.
,
George
T.
,
Hall
B.
,
Basiji
D.
,
Ortyn
W.
,
Morrissey
P.
.
2008
.
Quantitative image based apoptotic index measurement using multispectral imaging flow cytometry: a comparison with standard photometric methods.
Apoptosis
13
:
1054
1063
.
25
Mosmann
T. R.
,
Coffman
R. L.
.
1989
.
TH1 and TH2 cells: different patterns of lymphokine secretion lead to different functional properties.
Annu. Rev. Immunol.
7
:
145
173
.
26
Mathis
D.
,
Benoist
C.
.
2009
.
Aire.
Annu. Rev. Immunol.
27
:
287
312
.
27
Benlagha
K.
,
Park
S. H.
,
Guinamard
R.
,
Forestier
C.
,
Karlsson
L.
,
Chang
C. H.
,
Bendelac
A.
.
2004
.
Mechanisms governing B cell developmental defects in invariant chain-deficient mice.
J. Immunol.
172
:
2076
2083
.
28
Hinterberger
M.
,
Aichinger
M.
,
Prazeres da Costa
O.
,
Voehringer
D.
,
Hoffmann
R.
,
Klein
L.
.
2010
.
Autonomous role of medullary thymic epithelial cells in central CD4+ T cell tolerance.
Nat. Immunol.
11
:
512
519
.
29
de Bont
E. S.
,
Reilly
C. R.
,
Lo
D.
,
Glimcher
L. H.
,
Laufer
T. M.
.
1999
.
A minimal level of MHC class II expression is sufficient to abrogate autoreactivity.
Int. Immunol.
11
:
1295
1306
.
30
Azzam
H. S.
,
Grinberg
A.
,
Lui
K.
,
Shen
H.
,
Shores
E. W.
,
Love
P. E.
.
1998
.
CD5 expression is developmentally regulated by T cell receptor (TCR) signals and TCR avidity.
J. Exp. Med.
188
:
2301
2311
.
31
Moran
A. E.
,
Holzapfel
K. L.
,
Xing
Y.
,
Cunningham
N. R.
,
Maltzman
J. S.
,
Punt
J.
,
Hogquist
K. A.
.
2011
.
T cell receptor signal strength in Treg and iNKT cell development demonstrated by a novel fluorescent reporter mouse.
J. Exp. Med.
208
:
1279
1289
.
32
Nedjic
J.
,
Aichinger
M.
,
Emmerich
J.
,
Mizushima
N.
,
Klein
L.
.
2008
.
Autophagy in thymic epithelium shapes the T-cell repertoire and is essential for tolerance.
Nature
455
:
396
400
.
33
Lee
W. T.
,
Yin
X. M.
,
Vitetta
E. S.
.
1990
.
Functional and ontogenetic analysis of murine CD45Rhi and CD45Rlo CD4+ T cells.
J. Immunol.
144
:
3288
3295
.
34
Ostman
S.
,
Rask
C.
,
Wold
A. E.
,
Hultkrantz
S.
,
Telemo
E.
.
2006
.
Impaired regulatory T cell function in germ-free mice.
Eur. J. Immunol.
36
:
2336
2346
.
35
Lee
W. T.
,
Vitetta
E. S.
.
1991
.
Virgin T cells do not provide help for antigen-specific B cells in the absence of IL-4, IL-5, and IL-6.
Int. Immunol.
3
:
907
916
.
36
Swain
S. L.
,
Weinberg
A. D.
,
English
M.
.
1990
.
CD4+ T cell subsets. Lymphokine secretion of memory cells and of effector cells that develop from precursors in vitro.
J. Immunol.
144
:
1788
1799
.
37
Hosken
N. A.
,
Shibuya
K.
,
Heath
A. W.
,
Murphy
K. M.
,
O’Garra
A.
.
1995
.
The effect of antigen dose on CD4+ T helper cell phenotype development in a T cell receptor-alpha beta-transgenic model.
J. Exp. Med.
182
:
1579
1584
.
38
Seder
R. A.
,
Paul
W. E.
.
1994
.
Acquisition of lymphokine-producing phenotype by CD4+ T cells.
Annu. Rev. Immunol.
12
:
635
673
.
39
Dalai
S. K.
,
Khoruzhenko
S.
,
Drake
C. G.
,
Jie
C. C.
,
Sadegh-Nasseri
S.
.
2011
.
Resolution of infection promotes a state of dormancy and long survival of CD4 memory T cells.
Immunol. Cell Biol.
89
:
870
881
.
40
Sprent
J.
,
Surh
C. D.
.
2011
.
Normal T cell homeostasis: the conversion of naive cells into memory-phenotype cells.
Nat. Immunol.
12
:
478
484
.
41
Sadegh-Nasseri
S.
,
Dalai
S. K.
,
Korb Ferris
L. C.
,
Mirshahidi
S.
.
2010
.
Suboptimal engagement of the T-cell receptor by a variety of peptide-MHC ligands triggers T-cell anergy.
Immunology
129
:
1
7
.
42
Dalai
S. K.
,
Mirshahidi
S.
,
Morrot
A.
,
Zavala
F.
,
Sadegh-Nasseri
S.
.
2008
.
Anergy in memory CD4+ T cells is induced by B cells.
J. Immunol.
181
:
3221
3231
.
43
Mirshahidi
S.
,
Huang
C. T.
,
Sadegh-Nasseri
S.
.
2001
.
Anergy in peripheral memory CD4+ T cells induced by low avidity engagement of T cell receptor.
J. Exp. Med.
194
:
719
731
.
44
Jameson
S. C.
2002
.
Maintaining the norm: T-cell homeostasis.
Nat. Rev. Immunol.
2
:
547
556
.
45
Murali-Krishna
K.
,
Ahmed
R.
.
2000
.
Cutting edge: naive T cells masquerading as memory cells.
J. Immunol.
165
:
1733
1737
.
46
Sharma
D.
,
Kumar
S. S.
,
Checker
R.
,
Raghu
R.
,
Khanam
S.
,
Krishnan
S.
,
Sainis
K. B.
.
2009
.
Spatial distribution, kinetics, signaling and cytokine production during homeostasis driven proliferation of CD4+ T cells.
Mol. Immunol.
46
:
2403
2412
.

The authors have no financial conflicts of interest.