Class II transactivator (CIITA) is an unusual transcriptional coactivator in that it contains a functionally important, GTP-binding consensus domain. To assess the functional role of the GTP-binding domain of CIITA in vivo, we have generated knockout mice that bear a mutation in the CIITA gene spanning the GTP-binding domain. Upon analysis, these mice show no detectable CIITA mRNA; hence, they represent mice with deleted CIITA rather than mice with defects in the GTP-binding domain only. In these knockout mice, MHC class II expression is nearly eliminated, although a faint RT-PCR signal is visible in spleen, lymph node, and thymus, suggestive of the presence of CIITA-independent regulation of MHC class II expression. Invariant chain expression is also greatly reduced, but to a lesser extent than MHC class II. Serum IgM is not decreased, but the serum IgG level is greatly reduced, further confirming the absence of MHC class II Ag-dependent Ig class switching. Induction of MHC class II expression by IL-4 or LPS was absent on B cells, and Mac-1+ cells showed no detectable induction of MHC class II by either IL-4, LPS, or IFN-γ. These findings demonstrate a requirement for CIITA in IFN-γ-, IL-4-, and endotoxin-induced MHC class II expression as well as the possibility of rare CIITA-independent MHC class II expression.
The induction of an immune response is dependent upon the ability of T lymphocytes to recognize antigenic peptides. T lymphocytes recognize and respond to antigenic peptides only in association with MHC class I and II molecules. The MHC class II molecule presents antigenic peptides to CD4+ T cells through interactions with both the TCR and the CD4 molecule, leading to the activation of these cells. CD4+ Th cells subsequently synthesize cytokines that activate B cell or CD8+ cytotoxic T cell functions. MHC class II molecules are also crucial in the positive and negative thymic selection that is required to generate a self-tolerant yet MHC-restricted mature T cell repertoire.
The presentation of antigenic peptides to CD4+ T cells by MHC class II molecules requires the coexpression of invariant chain (Ii)5 and DM in humans or H2-M in mice. In the endoplasmic reticulum, newly synthesized MHC class II molecules associate with Ii and form nonameric complexes consisting of three MHC class II dimers and an Ii trimer (reviewed in Ref. 1). The targeting signal in Ii shuttles these complexes through the Golgi apparatus to a specialized endosomal/lysosomal compartment where class II-associated invariant chain peptide (CLIP) is replaced by a foreign antigenic peptide. The DM molecule plays a critical role in the removal of CLIP and the loading of foreign peptides (2, 3, 4, 5, 6, 7). Peptide-loaded MHC class II complexes subsequently move to the cell surface to present peptides to CD4+ T cells.
Unlike MHC class I molecules, which are expressed by all cell types, the expression of MHC class II molecules in mice is restricted to B cells, professional APCs, and thymic epithelial cells (8, 9). In most cell types, except B cells, the expression of MHC class II molecules is inducible by IFN-γ. This complex pattern of MHC class II gene expression is regulated by a number of transcription factors, including the transcriptional coactivator class II transactivator (CIITA). CIITA, which was first isolated by complementation cloning of the MHC class II-defective cell line RJ2.2.5, restores MHC class II expression in cells of the bare lymphocyte syndrome complementation group A (10). CIITA is expressed only in cells that express MHC class II molecules and is the IFN-γ inducible trans-acting factor required for the induction of MHC class II gene expression (11, 12, 13). CIITA is also necessary for expression of the other IFN-γ inducible genes involved in Ag presentation, namely DM and Ii (11, 14, 15, 16). Promoter analyses of the DM, Ii, and MHC class II genes have revealed the presence of several shared cis-acting regulatory elements known as the W, X, and Y motifs; however, CIITA does not appear to bind directly to these promoter sequences.
Structure function analyses of CIITA have been performed to understand the mode of action of CIITA. These analyses revealed that a conventional acidic domain consisting of residues 26–137 (17, 18) and a proline/serine/threonine-rich domain consisting of residues 163–322 are important for transcriptional activities (19). Intriguingly, a GTP-binding consensus sequence contained in residues 421–561 is also important for CIITA function (19, 20). The involvement of acidic and proline/serine/threonine-rich domains in transcriptional activation has been observed for several transcription factors (21, 22, 23, 24, 25); however, a role for a GTP-binding domain in transcriptional activation has not been described previously. The GTP-binding sequence is typically important in signal transduction, protein synthesis, and intracellular protein transport. To further assess the functional role of the GTP-binding domain of CIITA in vivo, we targeted a region of the CIITA gene that spans the GTP-binding domain for removal by homologous recombination. This study reports the production and characterization of such a knockout (KO) mouse strain and demonstrates that CIITA is critical for constitutive as well as inducible expression of MHC class II by a variety of biologic effector molecules. However, a very low level of MHC class II mRNA is still detectable by RT-PCR. This finding is in agreement with a recent report indicating that low levels of class II are still detected on dendritic cells from CIITA null mice (26).
Materials and Methods
Mouse CIITA genomic clones were isolated from a 129/Sv library (Stratagene, La Jolla, CA) and mapped using restriction enzymes and human CIITA cDNA probes (19). The targeting vector was constructed by replacing a 3.0-kb HindIII fragment containing the GTP-binding domain of CIITA with a neomycin (neo) gene cassette in plasmid pPNT (27), which contains the HSV thymidine kinase gene (Fig. 1,A). The targeting vector was linearized with NotI and electroporated into E14TG2a embryonic stem (ES) cells. The cells were placed under double selection with G418 and ganciclovir to enrich for those that had undergone homologous recombination. Clones were subjected to Southern blot analysis using a 1.2-kb KpnI CIITA gene fragment as the probe. Of the 127 clones screened, 3 carried the GTP-binding domain deletion. These were identified by the presence of a 7-kb hybridizing mutant band in addition to the 12-kb wild-type (WT) band that was seen in nontargeted cells (Fig. 1 B). Targeted ES cell clones were injected into C57BL/6 blastocysts, the resulting embryos were implanted into foster mothers, and these embryos subsequently gave rise to 18 chimeric mice. The male chimeras were mated to 6-wk-old female C57BL/6 mice obtained from The Jackson Laboratory (Bar Harbor, ME). Offspring carrying the targeted CIITA gene were identified by Southern blot analysis of tail DNA, which was prepared as described previously (28). Male and female heterozygotes were intercrossed to produce the deletion mutant (hereafter designated as CIITA−/−). All mice were propagated at the University of North Carolina Animal Facility (approved by the Institutional Animal Care and Use Committee) and were maintained in a pathogen-free colony.
Thymus, spleen, and lymph node (LN) RNAs were isolated from 5- to 6-wk-old mice using Trizol reagent (Molecular Research Center, Cincinnati, OH) according to the manufacturer’s protocol. For Northern blot analyses, total RNA was electrophoresed in a 1% formaldehyde agarose gel, transferred to Nytran membranes (Schleicher and Schuell, Keene, NH), and hybridized with random-primed genomic or cDNA probes according to the manufacturer’s protocol (Schleicher and Schuell). The probes used were: human β-actin as a control, pI-Aβ2 for MHC class II (a gift of Carolyn Doyle, Duke University, Durham, NC), pmIip34 for Ii (a gift of Ron Germain, National Institutes of Health, Bethesda, MD), and pH-2IIa for MHC class I (a gift of Sherman Weissman, Yale University, New Haven, CT). For RT-PCR, first-strand cDNA was synthesized using reverse transcriptase (Life Technologies, Grand Island, NY); PCR was performed using AmpliTaq DNA polymerase (Perkin-Elmer, Norwalk, CT) in a Perkin-Elmer 9600 thermocycler. PCR conditions were: predenaturation at 94°C for 20 s, followed by 30 cycles of denaturation at 94°C for 20 s, annealing at 70°C for 20 s, and elongation at 72°C for 30 s. The primers used were: H-2Aa, 5′-TATGTGGACTTGGATAAGAAG-3′ (sense), 5′-ACAAAGCAGATAAGGGTGTTG-3′ (antisense); H-2Dd, 5′-CCCTGACCTGGCAGTTGAATG-3′ (sense), 5′-AGCTCCAAGGATGACCACAGC-3′ (antisense); mouse β-actin, 5′-GGCATTGTTACCAACTGGGAC-3′ (sense), 5′-ACCAGAGGCATACAGGGACAG-3′ (antisense); mouse Ii, 5′-GTGTCTGTTTCATCGTCCCAG-3′ (sense), 5′-AAGGCAGCAAATGTGTCCAGC-3′ (antisense); mouse CIITA, 5′-TGCAGGCGACCAGGAGAGACA-3′ (sense), 5′-GAAGCTGGGCACCTCAAAGAT-3′ (antisense); and CIITA-5′, 5′-GCAGCTACCTGGAACTCCTTA-3′ (sense), 5′-CTCATTTACACGGGAGGTCAG-3′ (antisense).
Flow cytometric analyses
Cells from the thymus and spleen were dispersed in ice-cold PBS supplemented with 2% FBS (PBS-FBS). Spleen erythrocytes were lysed by hypotonic shock. Splenic cells were incubated with rat anti-mouse CD16/CD32 FcγIII/IIR Fc Block (PharMingen, San Diego, CA) at 4°C for 15 min to block Fc receptors before staining. The cells were resuspended at 1 × 108 cells/ml in PBS-FBS. A total of 10 μl of the cell suspension was incubated with Abs on ice for 30 min in a final volume of 100 μl. Cells that were incubated with biotinylated Abs were washed three times with PBS-FBS and subsequently stained with FITC or PE avidin D (Vector Laboratories, Burlingame, CA) on ice for 30 min. The stained cells were fixed in 2% paraformaldehyde and analyzed by FACScan (Becton Dickinson, Mountain View, CA). The Abs used were: biotinylated anti-H-2Kb/H-2Db for MHC class I, anti-I-Ab/I-Eb-FITC for MHC class II, anti-CD45R/B220-PE for B cells, anti-CD4-FITC for CD4+ T cells, anti-CD8-PE for CD8+ T cells, and biotinylated anti-CD11b for Mac-1. All Abs were purchased from PharMingen.
Thymus, LN, and splenic tissues were snap-frozen in liquid nitrogen for cryostat sectioning. Sections (4–5 μm thick) were air dried and fixed in acetone for 1 min. Before staining, sections were rehydrated in 1× PBS and subsequently incubated with primary Abs. The primary Abs (all purchased from American Type Culture Collection, Manassas, VA) used in this study were: GK1.5 for CD4 staining, Lyt2.2 for CD8 staining, M5/114.15.2 for MHC class II, and M1/220.127.116.11.HLK for MHC class I. The primary Abs were incubated for 30 min at room temperature followed by incubation with affinity-isolated, FITC-conjugated goat anti-rat IgG for 1 h at room temperature. IgG2b was used as an isotype control for MHC class II.
Serum IgM and IgG levels
Serum was isolated from blood using serum separators (Microtainer, Becton Dickinson) according to the manufacturer’s protocol. For ELISA, 100 μl of anti-mouse IgM or IgG (Sigma, St. Louis, MO) at 30 μg/ml were plated and incubated for 2 h at room temperature. The plate was washed three times with 200 μl of PBS/well and blocked overnight with 1% BSA (BSA/PBS) at 4°C. The plate was washed again as described and incubated with 100 μl of the diluted samples for 3 h at room temperature; the sample sera were diluted 1/100, 1/250, 1/500, 1/750, and 1/1000 with 0.02% BSA/PBS. The plate was washed, and alkaline phosphatase-conjugated anti-mouse IgG or anti-mouse IgM (Sigma) was added to the wells and incubated for 1 h at room temperature. After a final wash, 100 μl of phosphatase substrate (Sigma) was added to the wells; the OD results were read at 405 nm.
IL-4, LPS, and IFN-γ injections
An established protocol was used to demonstrate the IL-4 induction of MHC class II expression (29). Briefly, 6- to 8-wk-old mice were injected i.v. with 2 μg of IL-4 (Endogen, Woburn, MA) and 5 μg of anti-IL-4 Ab 11B11 (PharMingen) or with normal saline as a control. Anti-IL-4 Ab was added because of its potentiating effects on IL-4-induced class II expression, as documented previously (29). Mice were sacrificed after 48 h, and single-cell suspensions were prepared from spleens and LNs. Cells were stained with anti-I-Ab/I-Eb-FITC and anti-CD45R/B220-PE or with biotinylated anti-CD11b and subjected to flow cytometric analyses as described above. LPS and IFN-γ (Genzyme, Cambridge, MA) were injected i.p. at doses of 400 μg and 50,000 U, respectively. Again, spleen and LNs were harvested 48 h later, and cells were stained with anti-I-Ab/I-Eb-FITC and anti-CD45R/B220-PE or with biotinylated anti-CD11b and subjected to flow cytometric analyses.
Generation of mice lacking the GTP-binding domain of CIITA
The GTP-binding motifs in CIITA (amino acids 421–561) consist of the phosphate-binding, Mg2+-binding, and guanine-binding motifs (19). Mutation of any of these motifs greatly diminishes the ability of CIITA to trans-activate the MHC class II promoter (19, 20). To further assess the in vivo function of the GTP-binding domain of CIITA, a 3.0-kb HindIII fragment containing the GTP-binding domain of CIITA was replaced by a neomycin (neo) gene cassette to generate a mouse lacking the GTP-binding domain (CIITA−/−) (Fig. 1, A and B, also see Materials and Methods). Spleen, thymus, and LN RNAs were isolated from WT control mice and CIITA−/− mice and examined for CIITA gene expression by RT-PCR (Fig. 2). The use of PCR was necessary because endogenous CIITA levels are extremely low. The use of serially diluted cDNA in PCRs permits the semiquantitative comparison of two samples in the range where cDNA is not in excess. These experiments were repeated in three mice with identical results. In the WT control mice, CIITA expression was detected in the spleen, thymus (albeit to a lesser extent), and LNs. CIITA expression was not detected in these same cells isolated from the CIITA−/− mice using either oligonucleotides specific for the GTP-binding domain or for the acidic domain (Fig. 2, A and B). The acidic domain lies upstream of the targeted region.
Reduction of MHC class II and Ii gene expression in CIITA−/− mice
To assess whether the CIITA−/− phenotype affected MHC class II expression, cDNAs prepared from WT and CIITA−/− mice were examined for MHC class II and Ii gene expression by RT-PCR. As shown in Fig. 3,A, the expression of MHC class II was greatly reduced in the spleen, thymus, and LNs of the CIITA−/− mice compared with that of the WT controls, although the signal was not completely eliminated. The Ii transcript was also reduced, although to a lesser extent than MHC class II. There was no significant or consistent difference in MHC class I or β-actin gene expression in these mice. Northern blot analyses were also performed using total RNA isolated from the spleen, thymus, and LNs to verify the RT-PCR results. According to this assay, the Ii RNA was markedly reduced and MHC class II RNA was not visible; however, there were no changes in MHC class I and β-actin RNA levels in the CIITA−/− mice compared with the control mice (Fig. 3 B). The differences in gene expression between WT and CIITA−/− mice were more clearly revealed by Northern blot analyses than by PCR. This is expected, as PCR is likely to amplify minor signals, thus mitigating some of the actual differences between the WT and gene KO mice. Regardless, it is clear from both types of analyses that the targeted deletion results in a great change in MHC class II and Ii gene expression in vivo.
Cell surface expression of MHC class II molecules
Using flow cytometric analyses, the presence of cell surface MHC class I and II molecules was examined on splenic B cells isolated from CIITA−/− and WT mice (Fig. 4). Approximately 50–60% of the splenic cells were B cells, and the CIITA−/− mice showed a small increase in this population. In the CIITA−/− mice, virtually all B220+ B cells were MHC class II negative (99.5%). In the control mice, nearly all of B220+ B cells were MHC class II positive (92.3%). However, there was no significant difference in MHC class I expression on B220+ B cells between the CIITA−/− and control mice. The splenic Mac-1+ cells were also normal in number but devoid of MHC class II expression in the CIITA−/− mice, whereas in the control mice, 60% of the Mac-1+ cells were MHC class II positive (data not shown). Immunofluorescence studies were also performed on tissue sections, and the results were identical with the flow cytometric analyses (data not shown). Taken together, the RT-PCR, Northern blot, flow cytometry, and fluorescent immunohistology data clearly show that there is a great reduction of MHC class II expression in the CIITA−/− mice.
Reduction of CD4+ T cells and compensatory up-regulation of CD8+ T cells
In I-Aβ−/− (MHC class II gene KO) mice, a sharp reduction in the number of CD4+ T cells and a proportional increase in CD8+ T cells were observed in the spleen, thymus, and LNs (30). When splenic and thymic cells from CIITA−/− mice were stained with anti-CD4 and anti-CD8 mAbs and analyzed by flow cytometry, we also saw a great reduction of CD4+ T cells present only in both the thymus and spleen (Fig. 5 A). In the CIITA−/− mice, CD4+ T cells represent 1.8% of the cells in the thymus and 4.5% of the cells in the spleen, whereas in the WT controls, CD4+ T cells comprised ≤14.3% in the thymus and 21.6% of the cells in the spleen. There was a compensatory increase in CD8+ T cells (from 14.2% to 23.1%) in the spleens of the CIITA−/− mice. There was also a slight increase in CD4+ CD8+ T cells (from 79.4% to 92.5%) in the thymi of the CIITA−/− mice.
To visualize the T cell distribution, fluorescent immunohistology was performed on spleen, thymus, and LN sections with anti-CD4 and anti-CD8 mAbs. Fig. 5,B shows thymus and spleen sections from CIITA−/− and control mice. In thymic sections from the CIITA−/− mice, very few cells stained with anti-CD4; however, we saw a compensatory increase in cells stained with anti-CD8 in the medulla compared with the normal staining pattern seen in the WT mice (Fig. 5,B, left). Spleen sections from the CIITA−/− mice showed a lack of perivascular staining with anti-CD4, with proportionally increased staining with anti-CD8 (Fig. 5 B, right); WT mice showed a normal perivascular staining. The same results were obtained with LNs (data not shown).
Taken together, these data indicate that the differentiation of CD4+ T cells is greatly perturbed in the thymus in CIITA−/− mice due to the drastic reduction of MHC class II molecule expression. This leads to a decrease in the peripheral CD4+ T cell population and a compensatory up-regulation of CD8+ T cells in CIITA−/− mice.
Reduction of IgG but not IgM levels in CIITA−/− mice
Although the number of B220+ B cells in the spleen appears to be normal in the CIITA−/− mice, virtually all of these cells are MHC class II negative. When serum IgM and IgG levels were examined, there was a slight increase in serum IgM and a severe reduction in serum IgG levels in CIITA−/− mice (Fig. 6). This parallels the finding in I-Aβ−/− mice and shows that there is a defect in MHC class II-dependent Ig class-switching from IgM to IgG (30, 31).
Elimination of IL-4-, endotoxin-, and IFN-γ-induced MHC class II expression in CIITA−/− mice
CIITA is clearly an important molecule for both constitutive expression and the IFN-γ induction of MHC class II expression. Using the mice generated in this report, we confirmed the finding that the IFN-γ induction of MHC class II was eliminated (data not shown) (32). This was not surprising, because the role of CIITA in the IFN-γ induction of MHC class II is well documented (11, 12, 13). However, the role of CIITA in mediating the induction of MHC class II expression by other biologic inducers had not been elucidated previously. To address this, we tested the capacity of IL-4 and LPS (endotoxin) to induce MHC class II expression in the absence of CIITA. IL-4 and LPS are known to up-regulate MHC class II expression on both B lymphocytes and macrophages (33), whereas IFN-γ up-regulates expression on macrophages but not B cells. As expected, i.v. injection of IL-4 increased MHC class II expression on total splenic cells from WT mice (Fig. 7,A). When these cells were analyzed further, it was apparent that MHC class II was significantly amplified on B220+ B cells and Mac-1+ cells in the spleens of WT mice (Fig. 7,A). In contrast, MHC class II expression was negligible in cells isolated from CIITA−/− mice that had been treated with IL-4. The CIITA−/− mean channel fluorescence, although slightly above baseline, clearly shows an absence of induction by IL-4 when compared with saline-treated controls (Fig. 7,A). Only 0.41% of all CIITA−/− events fell outside of the MHC class II negative quadrants, and a similar number of “positive” events (0.43%) was seen with our isotype control (Fig. 7,B). Therefore, this low level of binding probably represents nonspecific staining of our Abs and not actual MHC class II expression. Treatment with LPS (Fig. 7 C) produced similar results. Identical results were obtained with the LNs of these animals (data not shown). These data indicate that CIITA expression is pivotal for controlling MHC class II induction by IL-4, endotoxin, and IFN-γ.
A CIITA−/− gene KO strain was produced by the targeted deletion of a genetic region that includes the GTP-binding domain of CIITA. The GTP-binding consensus motif was targeted because of previous studies showing an important role for this motif in the trans-activation function of CIITA (19, 20). The phenotype of this strain is similar to two other CIITA gene KO strains produced by the targeting of other sequences (26, 32). The targeted deletion reported here results in the elimination of detectable CIITA transcript, as judged by RT-PCR analysis using sequences in an acidic domain upstream of the targeted deletion. Using the mice generated in this study, we show that CIITA is required for both constitutive and induced expression of MHC class II on different subpopulations of lymphoid cells. Thus, CIITA is required for the regulation of MHC class II expression under all of the conditions examined in this report.
Three primary inducers were tested, and all three were shown to depend upon CIITA for the induction of MHC class II expression. IFN-γ is the most potent inducer of MHC class II, and many in vitro analyses have pointed to CIITA as the critical molecule in mediating the IFN-γ response; thus the involvement of CIITA in this response is not surprising (11, 12, 13). Composite studies from our group and other groups indicate that IFN-γ up-regulates CIITA promoter activation, which then leads to the downstream activation of MHC class II expression (34, 35). IFN-γ up-regulates the activity of two distinct CIITA promoters: promoter III is most likely activated directly by the IFN-γ-induced phosphorylation of the STAT1 protein, whereas promoter IV is activated primarily by interferon regulatory factor-1 (IRF-1) and, to a lesser extent, by the STAT1 protein (35). The IL-4 induction of MHC class II is less well studied, and the role of CIITA in this response has not been tested. Although IL-4 was known to induce the binding of DNA-binding proteins, which recognize DNA sequences in the MHC class II promoter by gel shift analysis, the involvement of these DNA-binding proteins in MHC class II expression was not directly verified (36). The results in this report indicate that IL-4 induction of MHC class II transcription is in fact mediated by CIITA, and this dependency upon CIITA is present in both B220+ B cells and Mac-1+ cells. This finding provides a starting point for revisiting the issue of MHC class II induction by IL-4. Finally, endotoxin has long been shown to induce MHC class II, and there is some evidence that endotoxin and IL-4 constitute overlapping pathways (37). Others have shown that IL-1, a primary product of endotoxin treatment, induces IL-4 (38). It is likely that both IL-4 and endotoxin activate MHC class II expression by enhancing the expression of CIITA.
These data clearly demonstrate that CIITA is a crucial regulator of MHC class II expression, with its elimination abrogating most of the MHC class II expression. However, RT-PCR analysis shows some residual MHC class II expression in all primary and secondary lymphoid tissues examined. Several attempts in our laboratory to identify the cell(s) harboring this CIITA-independent, MHC class II expression by immunohistochemistry and FACS have not generated credible data, because this cell is likely very rare. Previously, Chang et al. (26) detected MHC class II in the thymic tissue of their CIITA−/− mice (32). Since the preparation of this manuscript, Williams et al. have shown the presence of Ia expression among dendritic cells found in the LNs of CIITA knockouts, although the expression level was significantly reduced. Other reports have used mutants with defects in MHC class II expression to dissociate MHC class II expression from CIITA expression by in vitro analysis (39, 40, 41). All of these findings support a crucial role for CIITA-dependent MHC class II gene expression under a variety of conditions; however, one or more than one CIITA-independent pathway(s) also exist. The biologic function of both pathways will be of interest and testable with the mouse strain described here. Another consideration that is testable is the possibility that an isotype-specific control of MHC class II genes is differentially dependent upon CIITA (39, 42, 43).
The one difference between the CIITA−/− strain reported here and the CIITA−/− mouse strain produced by Chang et al. (32) is the level of Ii expression. The CIITA−/− mice described here exhibit a greater reduction of Ii gene expression, although this is likely attributed to inherent differences in the assays used to measure Ii RNA levels. The previous report used RT-PCR, whereas both PCR and Northern blot analyses were used in this report. The RT-PCR assay is more prone to artifacts, particularly in underestimating differences among samples. In this report, the Northern blot analysis showed a significant drop in the level of Ii gene expression, although the reduction is less pronounced than that for MHC class II transcripts. This is most likely due to the presence of additional regulatory elements in the Ii promoter, such as Sp-1, NF-κB, and an additional NF-Y binding site, all of which contribute to the up-regulation of Ii (44, 45, 46). This dichotomy of MHC class II and Ii gene expression has a biologic basis: in addition to its role in Ag processing, Ii has also been shown to regulate/direct B cell development (47).
Previously, we and others have demonstrated in cell lines that CIITA can regulate the cytokine-induced (e.g., IFN-γ) expression of MHC class I genes and Ags in vitro (48, 49). This regulation is most evident in cell lines that either do not express or express very low levels of basal MHC class I expression. In contrast, MHC class I expression is not affected by CIITA in cell lines that express moderate to high levels of MHC class I. The control of MHC class I promoters by CIITA is mediated through the α site in these promoters. The α site is similar but not identical with an AP-1 binding site, and its interaction with AP-1 is, at best, poor. However, the importance of the α site is substantiated by genomic footprint analysis, which shows that the in vivo occupancy of this site is correlated with MHC class I gene expression (50).
The creation of CIITA−/− mice allowed us to determine whether CIITA can regulate MHC class I expression in vivo. This report shows that MHC class I expression is not grossly altered in these mice; however, it should be emphasized that the tissues examined in this report typically express relatively high levels of MHC class I. The effect of CIITA in vitro is only evident in cell lines with little MHC class I expression (49). Thus, the in vivo effect of CIITA on MHC class I expression may be cell type-restricted; further analyses will be necessary to resolve the physiologic significance of MHC class I induction by CIITA.
In summary, a CIITA gene KO mouse strain was created by the targeted deletion of a region that includes the GTP-binding consensus sequence. The CIITA−/− mice exhibited a near elimination of constitutive MHC class II expression and induced expression in response to IL-4, LPS, and IFN-γ. Effects on MHC class II expression in this strain are correlated with a gross impairment of CD4+ T cell maturation in the thymus and with the depletion of these cells in the periphery. These mice will be useful as a vehicle for gene knock-in experiments as well as for the study of the biologic functions of CIITA and other molecules in the MHC class II Ag-processing pathway.
This work was supported by National Institutes of Health (NIH) Grants AI41580, AI29564, and AI41751, by National Multiple Sclerosis Society Grant RG1725 (to J.P.-Y.T.), by a National Multiple Sclerosis Society Fellowship (to J.F.P.), by a National Kidney Foundation Fellowship (to W.J.B.), by an NIH predoctoral fellowship (to N.J.F.), by NIH Grants HL46810 and HL52297 (to J.L.P.), and by NIH Grant DK 38108 (to B.H.K.).
Abbreviations used in this paper: Ii, invariant chain; CIITA, class II transactivator; WT, wild type; LN, lymph node; KO, knockout; ES, embryonic stem.