Myeloid Differentiation Factor 88 Is Required for Cross-Priming In Vivo Deborah Palliser, Hidde Ploegh and Marianne Boes This information is current as J Immunol 2004; 172:3415-3421; ; of September 24, 2021. doi: 10.4049/jimmunol.172.6.3415 http://www.jimmunol.org/content/172/6/3415 Downloaded from References This article cites 51 articles, 22 of which you can access for free at: http://www.jimmunol.org/content/172/6/3415.full#ref-list-1

Why The JI? Submit online. http://www.jimmunol.org/ • Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average by guest on September 24, 2021 Subscription Information about subscribing to The Journal of is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2004 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Myeloid Differentiation Factor 88 Is Required for Cross-Priming In Vivo1

Deborah Palliser,* Hidde Ploegh,2† and Marianne Boes†

We describe a role for myeloid differentiation factor 88 (MyD88) in the induction of functional CTLs in vivo, in response to exogenously administered Ag, using a heat shock fusion protein, hsp65-P1, as a model Ag. CD8 T cells transferred into MyD88- deficient animals produce normal numbers of CD8 effector cells that have normal activation marker profiles after immunization with hsp65-P1. However, these CD8 T cells produced significantly less IFN-␥ and showed reduced killing activity. This reduction in activation of functional CTLs appears to be unrelated to Toll-like receptor 4 function, because in vitro hsp65-P1-experienced Toll-like receptor 4-deficient dendritic cells (DCs), but not MyD88-deficient DCs, activated CD8 T cells to a similar extent to wild-type DCs. We identify a cross-presentation defect in MyD88-deficient DCs that, when treated with hsp65-P1 fusion protein,

results in surface display of fewer SIYRYYGL/class I MHC complexes. Thus, MyD88 plays a role in the developmental maturation Downloaded from of DCs that allows them to prime CD8 T cells through cross-presentation. The Journal of Immunology, 2004, 172: 3415–3421.

xogenous administration of Ag usually results in presen- opmental regulation plays a role in cross-presentation is unknown. tation to CD4 T cells via class II MHC, but class I MHC- Activation of a TLR allows recruitment of adapter proteins such as E restricted presentation of exogenously provided protein myeloid differentiation factor 88 (MyD88), Toll/IL-1R domain can occur as well, thereby inducing naive CTL to differentiate into containing adaptor protein, and possibly others. The binding of http://www.jimmunol.org/ activated CTL (1). This phenomenon is referred to as cross-pre- these adaptor proteins to Toll/IL-1R domains, present in the cyto- sentation, believed to play an important role in the protection plasmic region of the TLR, confers further specificity to the down- against tumor Ags and in antiviral immunity (2, 3). stream signaling events triggered by individual TLRs (11, 12). Protein Ags usually do not cross lipid membranes spontane- These events then ultimately lead to activation and nuclear trans- ously, and therefore fail to access the cytosolic class I MHC-pro- location of NF-␬B and activation of the DC. cessing pathway to elicit a CD8 T cell response. An exception are DCs that lack MyD88 are defective in the induction of cytokine fusion proteins composed of a mycobacterial heat shock protein genes upon treatment with the TLR4 ligand LPS, but their func- (hsp)3 coupled with protein fragments of varying length, which tional maturation appeared normal (13, 14). Signaling via TLR9 is readily stimulate the production of CD8 CTLs even in the absence completely abolished when MyD88 is absent (15). In addition, by guest on September 24, 2021 of adjuvants (4). Several such hsp fusion constructs elicit CD8 T MyD88 drives the initial priming of immune responses, particu- cell responses, and there may therefore be a unifying mechanism larly of the Th1 type (16). Earlier results from in vitro studies for their immunogenicity in the absence of added adjuvants (4–6). demonstrate the ability of DCs to adjust endocytosis, proteolysis, Under circumstances in which CD4 T cells and adjuvants are re- and display of class II MHC molecules in response to an inflam- quired for CD8 T cell responses, it is likely that these fusion pro- matory stimulus (17, 18). The class I MHC products also show teins function by activating dendritic cells (DCs) (7–9). In the ab- redistribution during the course of activation (19), although CD40 sence of adjuvant, treatment with hsp65 fusion proteins alone does ligation may be required. These observations all suggest a connec- not induce strong activation of DCs, although some maturation tion between activation of DCs upon receipt of a TLR-mediated does occur (6). What are the maturational requirements for a DC signal, and the ability of the DC to engage in Ag presentation. that allows cross-presentation of hsp fusion partners? Therefore, we asked whether there is a requirement for devel- DCs activated through ligation of Toll-like receptors (TLRs) opmental maturation of DCs for cross-presentation of peptides de- up-regulate cell surface expression of class I and II MHC, costimu- rived from exogenously added hsp. We show that lack of MyD88 latory molecules, and adhesion molecules (10), but whether devel- adapter protein results in suboptimal activation of functional CTL effectors from naive 2C transgenic T cells in vivo, when MyD88- deficient animals are immunized with hsp65-P1, which contains *Center for Cancer Research, Massachusetts Institute of Technology, Cambridge, MA b 02139; and †Department of Pathology, Harvard Medical School, Boston, MA 02115 the epitope for the H-2K -restricted 2C TCR-binding peptide SIYRYYGL (20). We show that the absence of MyD88 results in Received for publication September 23, 2003. Accepted for publication January 5, 2004. the formation of fewer MHC I-peptide complexes derived from the The costs of publication of this article were defrayed in part by the payment of page hsp65-P1 fusion protein on the surface of DCs. Therefore, we pro- charges. This article must therefore be hereby marked advertisement in accordance pose that this decrease may contribute to the observed decrease in with 18 U.S.C. Section 1734 solely to indicate this fact. CTL activation in vivo. Interestingly, this role for MyD88 in cross- 1 This work was supported by grants from the National Institutes of Health (to H.P.). priming appears to be independent of TLR4. D.P. was supported by a fellowship from StressGen Biotechnologies. M.B. was sup- ported by a fellowship from the Cancer Research Institute. 2 Address correspondence and reprint requests to Dr. Hidde Ploegh, Department of Pathology, Harvard Medical School, 200 Longwood Avenue, Boston, MA 02115. Materials and Methods E-mail address: [email protected] hsp65-P1 3 Abbreviations used in this paper: hsp, heat shock protein; DC, dendritic cell; MyD88, myeloid differentiation factor 88; PI, propidium iodide; TLR, Toll-like re- hsp65-P1 (StressGen Biotechnologies, Victoria, British Columbia, Can- ceptor; WT, wild type. ada) was prepared by linking a polypeptide called P1 to the C terminus of

Copyright © 2004 by The American Association of Immunologists, Inc. 0022-1767/04/$02.00 3416 ROLE FOR MyD88 IN CROSS-PRIMING mycobacterial (bacillus Calmette-Guerin) hsp65 (StressGen Biotechnolo- heat-inactivated FCS, 2 mM L-glutamine, 10 mM HEPES, 50 ␮M 2-ME, gies). P1, 26 aa in length, contains the SIYRYYGL octapeptide sequence penicillin, and streptomycin) containing 20 ng/ml murine GM-CSF (R&D flanked at its C-terminal side by 6 aa from ␣-ketoglutaraldehyde dehydro- Systems, Minneapolis, MN). The medium was replaced on days 2 and 4, genase and at its N-terminal side by 9 aa from OVA. The specific lysis of and on day 6 the cells (immature DCs) were harvested. The DCs were Kbϩ cells transfected with hsp65-P1 or P1 by a CTL clone, whose TCR isolated by positive sorting (MACS) using anti-CD11c Ab (purity Ͼ95%), binds specifically to the SIYRYYGL-Kb complex, indicated that the conducted according to the manufacturer’s instructions (Miltenyi Biotec). SIYRYYGL sequence can be excised intracellularly and presented with Kb The clone L3.100 was derived from 2C TCR transgenic mice (25). It was (6). The production of this hsp65-P1 in Escherichia coli and its purification maintained by weekly stimulation with irradiated P815 (Ldϩ) cells. The have been described (6, 21). After passage through a 0.2-␮m filter, the clone was used in assays 6–9 days following stimulation. protein was stored in sterile solution in PBS at 4°C or frozen at Ϫ80°C. The level of LPS in this preparation of hsp65-P1 was 6 EU/mg (by Limulus In vitro Ag presentation assay amebocyte lysate assay; Cape Cod Associates, Falmouth, MA). P1 and Purified DCs derived from bone marrow were plated at 4 ϫ 105 cells/well SIYRYYGL peptides were synthesized by the Massachusetts Institute of of a 96-well plate in RPMI 1640 medium. Protein or peptide was added at Technology Biopolymers Laboratory. Molecular weights were confirmed appropriate concentrations to each well. After 2-h incubation at 37°C, cells by amino acid analyses (Massachusetts Institute of Technology Biopolymer were fixed with 0.5% paraformaldehyde for 15 min at room temperature. Laboratory). The cells were then washed extensively with complete medium. A total of Mice 105 2C cells, either mouse lymph node or the CTL clone L3.100, was then added to each well, and the plate was incubated for 18 h, 37°C. We then C57BL/6 and C57BL/10 mice were purchased from The Jackson Labora- assessed the activation of 2C T cells either by FACS (CD69 expression), tory (Bar Harbor, ME), C57BL/10/ScNCr mice were from the National or by ELISA (IFN-␥ production). Cancer Institute (National Institutes of Health). MyD88Ϫ/Ϫ mice were kindly provided by S. Akira (Department of Host Defense, Research In- IFN-␥ ELISA Downloaded from stitute for Microbial Diseases, Osaka University, Osaka, Japan; Ref. 22). ␥ The 2C transgenic animals, expressing a TCR specific for SIYRYYGL-Kb, IFN- was measured using a kit from eBioscience (San Diego, CA), ac- were on a recombination-activating gene-1-deficient background (20). All cording to manufacturer’s specifications. Briefly, supernatants of DC/2C T cell cultures were collected after 18 h of cocultures and stored at Ϫ80°C. mice were maintained in a specific pathogen-free mouse facility. Studies ␥ were performed according to institutional guidelines for animal use Plates were coated with purified anti-mouse IFN- (clone XMG1.2). After and care. blocking with normal mouse serum, undiluted supernatants were incubated for 2 h (room temperature), and the assay was developed using a biotin- conjugated anti-mouse IFN-␥ (clone R4-6A2) and streptavidin. Immunizations http://www.jimmunol.org/ Naive 2C T cells were isolated from spleen and lymph nodes of 2C/re- Cytolytic T cell assays combination-activating gene mice. The cells were depleted of macrophages 51 b and DCs by magnetic sorting, according to the manufacturer’s instructions Cr-labeled T2-K cells were used as target cells. They were incubated (Miltenyi Biotec, Auburn, CA), using anti-CD11c- and anti-CD11b-coated with effector cells derived from fusion protein-injected mice for4hinthe ϩ ϩ ␮ microbeads. Purity was in the range of 80–85% (1B2 CD8 cells), as presence or absence of SIYRYYGL (1 M). Specific lysis was calculated Ϫ Ϫ judged by FACS analysis using a clonotypic Ab (1B2) specific for the 2C as follows: ((experimental counts spontaneous counts)/(total counts ϫ ␣␤ TCR. A total of 5 ϫ 106 2C cells was injected retro-orbitally. Three spontaneous counts)) 100. Control wells, from which the SIYRYYGL peptide was omitted, had background specific lysis of Ͻ10%; the back- days later, mice were injected s.c. in the scruff of the neck and base of the b tail with 50 ␮g of hsp65-P1, or equimolar amounts of unmodified hsp65 ground values were subtracted from the results shown. T2-K cells were a generous gift from P. Cresswell, Yale University School of Medicine (New plus P1 peptide (Massachusetts Institute of Technology Biopolymer Lab), by guest on September 24, 2021 in PBS. After 3 days, the frequency of 2C cells (percentage of total pro- Haven, CT). pidium iodide (PI)-negative cells that were 1B2ϩ), CD44 expression, and their ability to secrete IFN-␥ were determined by flow cytometry. These Results cells were also tested for cytolytic activity (see below). Hsp65 fusion proteins induce a CTL response in mice without the Flow cytometry need for adjuvant (6). The fusion protein used in this study consists of polypeptide P1 fused to the C terminus of mycobacterial hsp65. Fluorochrome-conjugated Abs to CD8␣, CD44, CD69, CD11c, CD38, b The P1 segment itself is a composite of OVA (OVA 251–257) and ICAM-1, I-A , and B7.2 were purchased from BD Biosciences (San Jose, ␣ CA). Y3 cross-reacts with MHC class I from H-2b (Kb) and H-2k haplo- -ketoglutaraldehyde dehydrogenase (6). The polypeptide P1 in- types and was affinity purified from culture supernatants from the Y3 hy- cludes the octapeptide SIYRYYGL, which when complexed with bridoma (obtained from American Type Culture Collection, Manassas, H2-Kb is a strong agonist for the TCR of 2C T cells (20). The VA) and labeled with fluorescein using FITC (Pierce, Rockford, IL). The maturation requirements of APCs to generate peptides from the clonotypic Ab 1B2, specific for the 2C TCR (23), was isolated from su- pernatants of cultured 1B2 cells, purified by protein G affinity chromatog- endocytosed fusion protein and their display on the cell surface as raphy, and conjugated to FITC. In addition to the specific Abs, purified a class I MHC-peptide complex are not fully understood. To study anti-FcR Ab (anti-CD16; BD Biosciences) was added to each sample to the role of DC maturation in the generation of CTL epitopes, we minimize nonspecific Fc-mediated binding to DCs. Cells were stained in relied on this well-defined in vivo model of cross-presentation of PBS containing 0.5% BSA and 0.1% NaN , and at least 20,000 live cells 3 hsp (6), in mice deficient in a major TLR adaptor protein, MyD88. (PI negative) were acquired on a FACSCalibur (BD Biosciences). Analysis was conducted using CellQuest software (BD Biosciences). Binding of We analyzed MyD88-deficient and wild-type (WT) mice for hsp65-P1 to DCs was measured by incubation with FITC-conjugated their ability to cross-present hsp65-P1 fusion protein in vivo. CD8 hsp65-P1 at 4°C in PBS containing 0.5% BSA and 0.1% NaN3 for 30 min. T cells were purified from lymph node cells of 2C transgenic mice Ͼ IFN-␥ assays ( 98% purity) to exclude the transfer of APCs from 2C mice, and were adoptively transferred into MyD88Ϫ/Ϫ or WT mice (5 ϫ 106 ␥ ϫ 6 IFN- was measured using a kit from Miltenyi Biotec. Briefly, 1 10 cells, by i.v. injection). Three days later, mice were immunized cells were restimulated with PMA (25 ng/ml) and ionomycin (500 ng/ml) ϩ for 2 h. After the cells were washed, they were resuspended in cold me- with either hsp65 P1 peptide or hsp65-P1 fusion protein (s.c. dium containing mouse IFN-␥ catch reagent on ice. Following an addi- immunization, 50 ␮g), or were left untreated. After another 3 days, tional 45-min incubation at 37°C, cells were stained with mouse IFN-␥ draining lymph nodes were isolated and analyzed for 2C T cell detection reagent, in addition to 1B2 and anti-CD8␣ (BD PharMingen, San expansion and activation (Fig. 1, A and B). The 2C T cell effector Diego, CA), and analyzed by FACS, gating on PI-negative cells. function was determined by IFN-␥ secretion by 2C T cells as well Cell preparation and culture as their ability to kill appropriate target cells by CTL assay (Fig. 1, DCs were generated, as described (24). In brief, bone marrow was flushed C and D). from the femur and tibia, RBC were lysed, and the remaining cells were For both WT and MyD88-deficient mice, the mixture of hsp65 cultured at 106 cells/ml in RPMI 1640 medium (supplemented with 10% and P1 (hsp65 ϩ P1) failed to evoke expansion of the transferred The Journal of Immunology 3417

expand in lymph nodes or acquire an Ag-experienced surface phenotype. The ability of naive precursors to differentiate into CTL effec- tors was analyzed further by IFN-␥ secretion, a cytokine produced by effector CD8 T cells. Immunization with hsp65-P1 induced high levels of IFN␥ by 2C T cells that had been transferred into WT mice. In contrast, immunization with hsp65-P1 in MyD88- deficient mice resulted in less IFN-␥ production by transferred 2C T cells ( p Ͻ 0.007). As expected, the mixture of hsp65 ϩ P1 failed to induce IFN-␥ in either strain of mice (Fig. 1C). When assayed for cytolytic activity, 2C T cells transferred into WT animals showed significantly higher levels of cytolytic activity than 2C T cells transferred into MyD88-deficient recipients (Fig. 1D). There- fore, immunization with hsp65-P1 results in cross-presentation of SIYRYYGL. This cross-presentation leads to proliferation and CD44 up-regulation by 2C T cells at comparable levels between WT and MyD88-deficient mice. However, in the MyD88-deficient mouse, this level or mode of cross-presentation is not sufficient for the generation of fully differentiated CTL. Downloaded from DCs are more effective than macrophages at presenting pro- cessed hsp fusion protein to naive CD8 2C T cells (6). We there- fore focused on the potential role of MyD88 in cross-presentation by DCs. Compared with WT, cytokine production by DCs is al- tered when MyD88-deficient DCs are stimulated with various

pathogens (16). Many different molecules are required to ensure http://www.jimmunol.org/ optimal T cell activation in response to a cross-presented Ag. What are the possible factors that play a role in cross-presentation, other FIGURE 1. Adoptive transfer of naive 2C T cells into MyD88-deficient than DC cytokine profiles, that are affected by absence of MyD88? ϫ 6 mice results in defective CTL activation. A total of 5 10 purified naive We first examined whether binding and uptake of hsp65-P1 by 2C cells was adoptively transferred by retro-orbital injection into either B6 Ϫ Ϫ DCs are influenced by the absence of MyD88. For all the in vitro or MyD88 / recipients. Three days later, mice were injected with either nothing, a mix of hsp65 plus P1, or hsp65-P1. All injections were admin- assays, we used bone marrow-derived DCs. Although bone mar- istered in saline, without added adjuvants. After an additional 3 days, drain- row-derived DCs may not be fully equivalent to the DCs involved ing lymph nodes were removed and analyzed for the following: A, expansion in cross-priming in vivo, we used them, as they are a homogenous by guest on September 24, 2021 of transferred 2C T cells; B, expression of CD44; C, IFN-␥ production; D, population of highly purified DCs. DCs purified from mouse tis- CTL killing ability. Values of p by unpaired two-tailed t test comparing B6 and sues tend to consist of a highly variable collection of DC popula- Indicates statis- tions that also vary in activation status. In addition, it is currently ,ء .MyD88-deficient mice injected with hsp65-P1 are shown tically significant difference. Results shown are representative of four experi- unknown which DC population is responsible for cross-priming in ments, with each group containing three to five mice. vivo, and various populations have been implicated (for review, see Ref. 26). FITC-conjugated hsp65-P1 was added to DCs cul- tured from bone marrow of MyD88-deficient and WT mice (40 2C T cells in the draining lymph nodes, and no up-regulation of the min, on ice). Equivalent levels of binding of labeled hsp65-P1 by activation marker CD44 was seen. In contrast, 2C cells transferred WT and MyD88-deficient DCs were observed (Fig. 2A). Uptake of into MyD88-deficient or WT hosts immunized with the fusion pro- labeled hsp65-P1 (12.5 ␮g/ml) was also measured (30 min, 37°C), tein hsp65-P1 showed robust expansion and an Ag-experienced and occurred with similar or more rapid kinetics by MyD88-defi- phenotype, characterized by high levels of CD44 (Fig. 1, A and B). cient DCs compared with WT DCs (Fig. 2B). We conclude that the Thus, the MyD88 gene product is not required for 2C T cells to absence of MyD88 does not impair the ability of DCs to bind and

FIGURE 2. Equivalent binding and endocytosis of hsp65-P1 in WT and MyD88-deficient DCs. A,B6and MyD88-deficient DCs were incubated at 4°C for 40 min in the presence of NaN3 with various concentra- tions of hsp65-P1. Cells were washed, and extent of binding was determined by FACS analysis. B,B6and MyD88-deficient DCs were incubated at 37°C for 30 min in the continuous presence of hsp65-P1 (12.5 ␮g/ ml, in medium in the absence of NaN3). 3418 ROLE FOR MyD88 IN CROSS-PRIMING endocytose the hsp65-P1 fusion protein. We have observed that DCs from WT and MyD88Ϫ/Ϫ mice are also equally efficient at endocytosis of fluorescein-labeled OVA (27). As binding and endocytosis of hsp65-P1 by WT and MyD88- deficient DCs were similar, we examined whether Ag presentation of hsp65-P1-derived SIYRYYGL on class I MHC is affected. DCs were allowed to endocytose hsp65-P1 at various concentrations for 2 h and were fixed, and naive 2C lymph node cells were added (overnight, 37°C). Activation of 2C T cells was then assessed by expression of the early activation marker CD69 (Fig. 3A). MyD88- deficient DCs activated fewer 2C T cells, as measured by up-reg- ulation of CD69. DCs lacking TLR4Ϫ/Ϫ obtained from C57BL/ 10ScNCr mice (28) show no such reduction compared with WT controls (C57BL/10) (Fig. 3B). These results suggest a role for MyD88 in cross-presentation of SIYRYYGL/class I MHC unre- lated to the function of TLR4. Although these results suggest that the difference in 2C activation observed is due to the formation of fewer peptide/MHC I complexes on the surface of MyD88-defi- cient DC, the level of peptide/MHC I complex formation is not Downloaded from being measured directly. To address whether fewer peptide/class I complexes are dis- played on MyD88-deficient DCs, we repeated the assay using both naive 2C cells and a 2C CTL clone, L3.100, which requires only TCR ligation to elicit a response (20). In addition, we titrated in

SIYRYYGL peptide to determine whether MHC I levels on WT http://www.jimmunol.org/ and MyD88-deficient DC were comparable. Both hsp65-P1 and exogenous SIYRYYGL peptide were analyzed for their ability to activate these naive and CTL effector 2C T cells. Exogenous SIYRYYGL peptide was added to MyD88-deficient and WT DCs (5 ϫ 10Ϫ9–10Ϫ11M) for 2 h, then fixed (as above), and either naive 2C or 2C CTL L3.100 cells were added (18-h culture, 37°C). The 2C T cells, either naive or L3.100, were acti- vated in a comparable manner by WT and MyD88-deficient DCs (Fig. 4A). To activate an equivalent number of the 2C CTL L3.100 by guest on September 24, 2021 when compared with naive 2C cells, 5- to 10-fold less SIYRYYGL peptide is required, demonstrating the increased sensitivity of the 2C L3.100 CTL clone (Fig. 4A). In addition, IFN-␥ production was measured from the supernatants of both naive 2C and 2C CTL L3.100 cells (Fig. 4C). WT and MyD88-deficient DCs were com- parable in their ability to activate 2C CTL clone cells in the pres- ence of exogenously added SIYRYYGL peptide. Naive 2C T cells

FIGURE 4. Activation of costimulation-dependent naive 2C T cells and FIGURE 3. SIYRYYGL/Class I complex formation is impaired in DCs a costimulation-independent 2C CTL clone is impaired upon incubation derived from MyD88-deficient mice, but not in DCs from TLR4-deficient with hsp65-P1-treated MyD88-deficient DCs. The assay was set up as in mice. A and B, For Ag presentation assays, 4 ϫ 105 DCs were incubated Fig. 3. For A and C, SIYRYYGL was added to DCs for 2 h before fixation. with various concentrations of hsp65-P1. After2hofincubation, the cells For B and D, hsp65-P1 was added to DCs for 2 h before fixation. Upper were fixed and washed extensively. Naive 2C cells were added and 18 h panel, SIYRYYGL; lower panel, hsp65-P1. A and B, CD69 up-regulation; later assayed for activation using the cell surface marker CD69 (A,B6and C and D, IFN-␥ production. MyD88Ϫ/Ϫ DC; B, B10 and ScNCr DC). Only PI-negative cells were an- alyzed. Experiments are representative of three. The Journal of Immunology 3419 were not fully activated by SIYRYYGL addition, as demonstrated in the display of SIYRYYGL/class I complexes at the surface of by decreased ability to produce IFN-␥ (Fig. 4C). The low level of DCs after uptake of hsp65-P1 fusion protein. Whether MyD88 is IFN-␥ production by naive 2C T cells may reflect the fact that involved in the generation of peptide/class I complexes, their in- naive T cells require the ligation of costimulatory molecules for tracellular transport to the cell surface, or localization in special- complete activation, which in the absence of an inflammatory stim- ized membrane domains (30) remains to be solved. ulus may not be sufficiently expressed on the DCs (29). There was no significant difference between MyD88-deficient and WT DCs to activate 2C T cells when treated with exogenously added Discussion SIYRYYGL peptide. We conclude that MyD88 is unlikely to af- Maturation of DCs in vitro is a process that modulates presentation fect the display of peptide-receptive class I MHC molecules at the of peptides on class I and II MHC at different levels, depending on cell surface. the stimuli received by the DC (17–19). Class II MHC molecules The ability to restimulate 2C cells by MyD88-deficient DCs is acquire their peptide cargo in endosomal compartments (31, 32), normal when exogenous SIYRYYGL peptide is provided, but is and endosomal transport is modulated by MyD88-dependent sig- MyD88 required to generate SIYRYYGL/class I complexes de- nals (27). In this study, we examined the requirement for MyD88 rived from hsp65-P1 fusion protein? The activation of naive 2C in H2-Kb-restricted cross-presentation of SIYRYYGL peptide de- and 2C CTL L3.100 T cells was measured after uptake of rived from hsp65-P1. Injection of hsp65-P1 resulted in similar lev- hsp65-P1 (20, 10, or 5 ␮g/ml, 18 h at 37°C). Activation of both els of SIYRYYGL-specific expansion of adoptively transferred 2C naive 2C T cells and 2C CTL L3.100 T cells was impaired, as T cells (on C57BL/6 background) and similar levels of CD44 sur- assessed by their ability to up-regulate the activation marker CD69 face expression in WT and MyD88-deficient recipients (Fig. 1, A Downloaded from (Fig. 4B). We also determined the level of secreted IFN-␥. The 2C and B). However, in MyD88-deficient animals, effector functions CTL L3.100 T cells that were activated by WT DCs produced of these CD8 T cells were impaired, as assessed by their ability to more IFN-␥ than 2C T cells activated by MyD88-deficient DCs produce IFN-␥ and acquire cytolytic capacity (Fig. 1, C and D). (Fig. 4D). Thus, the activation of both the naive 2C T cells and the To address how MyD88 is involved in cross-presentation, bone 2C CTL L3.100 via cross-presentation of SIYRYYGL derived marrow-derived DCs were treated with SIYRYYGL peptide or

from hsp65-P1 fusion protein is less effective in the absence of hsp65-P1, followed by incubation with 2C T cells (either naive or http://www.jimmunol.org/ MyD88. a 2C CTL clone). Compared with activation of naive CD8 T cells, As mentioned above, the impaired activation of naive 2C T cells the CTL clone was more readily activated due to its lack of re- (Figs. 3 and 4) could be explained by a decrease in expression of quirement for costimulation. Indeed, naive 2C T cells required at costimulatory molecules on the MyD88-deficient DCs. We there- least 5-fold more added peptide for activation, and produced little fore analyzed the expression of known costimulatory molecules on IFN-␥ compared with cloned 2C CTL cells. MyD88-deficient DCs WT and MyD88-deficient DCs after uptake of hsp65-P1 fusion were no less competent than WT DCs in activating 2C T cells protein. Upon activation of DCs, the expression of a number of when exogenous peptide was added, thereby excluding a role for activation markers, such as ICAM-1, I-Ab,Kb, CD38, and CD86, MyD88 in transport of class I MHC molecules to the cell surface is increased. When we compared DCs from WT animals with DCs (Fig. 4, A and C). Incubation of MyD88-deficient DCs with by guest on September 24, 2021 from MyD88-deficient animals, the relative increase in expression hsp65-P1 resulted in decreased activation of naive 2C T cells and of these markers was comparable (Fig. 5). Thus, MyD88 is not cloned 2C CTLs when compared with WT DCs (Fig. 4B). The involved in the cell surface expression of these costimulatory mol- production of IFN-␥ was also diminished in the 2C CTL clone ecules for cross-presentation of SIYRYYGL from hsp65-P1 fusion after culture with MyD88-deficient DCs when compared with WT. protein. It cannot be ruled out that other, as yet unidentified cell Naive 2C T cells did not produce any IFN-␥ after culture with surface molecules could show differences in expression and thus hsp65-P1-loaded WT or MyD88-deficient DCs. The number of play a role. cell surface SIYRYYGL/class I complexes is thus much reduced DCs can adjust their Ag presentation machinery during matu- in hsp65-P1-treated MyD88-deficient DCs compared with WT ration, as was shown using in vitro culture systems (17–19). We in DCs. Previous studies show that, in vivo, CD4 T cell priming is this study show, in an in vivo model, that MyD88 signals are affected in MyD88-deficient mice (33). Therefore, although we required for cross-presentation. Our data indicate a role for MyD88 have shown that surface display of SIYRYYGL/class I complex is affected by the absence of MyD88 in vitro, we cannot rule out that additional factors, such as altered CD4 T cell activation, may be a contributing factor in CD8 T cell priming in vivo. MyD88 functions as an adapter molecule for several TLRs be- sides TLR4. Therefore, the fusion protein hsp65-P1 may ligate TLRs other than TLR4. Such a scenario would explain the more severe phenotype of MyD88-deficient DCs when analyzed for cross-presentation when compared with TLR4-deficient DCs. The more severe defect could also be explained by a role for MyD88 in DC development. Indeed, MyD88 was originally described as a differentiation factor in myeloid cell development (34). We have excluded LPS contamination of the hsp65-P1 fusion protein as a cause for activation, as our preparations contain less than 6 EU/mg (see Materials and Methods), far below the minimally stimulatory FIGURE 5. Comparable levels of hsp65-P1 induced up-regulation of various costimulatory molecules in WT and MyD88-deficient DCs. A total concentration of LPS. of 4 ϫ 105 DCs was incubated for 18–24 h in the presence of 100 ␮g/ml Two pathways have been described for TLR4-mediated signal- hsp65-P1. The expression levels of proteins were assessed by FACS anal- ing; one requires the MyD88 adapter protein, whereas the other ysis gating on PI-negative cells, compared with the levels in the absence of does not. Specifically, the MyD88-dependent pathway is required hsp65-P1. for cytokine production by DCs and subsequent T cell activation, 3420 ROLE FOR MyD88 IN CROSS-PRIMING whereas the MyD88-independent pathway is involved in up-reg- 9. Schoenberger, S. P., R. E. Toes, E. I. van der Voort, R. Offringa, and C. J. Melief. ulation of costimulatory molecules and activation of the transcrip- 1998. T-cell help for cytotoxic T is mediated by CD40-CD40L interactions. Nature 393:480. tion factor IFN regulatory factor-3 (13, 35, 36). Earlier studies in 10. Barton, G. M., and R. Medzhitov. 2002. Control of adaptive immune responses MyD88-deficient mice have demonstrated a defect in CD4 T cell by Toll-like receptors. Curr. Opin. Immunol. 14:380. 11. Horng, T., G. M. Barton, and R. Medzhitov. 2001. TIRAP: an adapter molecule activation upon immunization as well as a skewing from Th1 to in the Toll signaling pathway. Nat. Immun. 2:835. Th2 cytokine production, thereby linking the lack of MyD88 to an 12. Takeuchi, O., and S. Akira. 2002. Genetic approaches to the study of Toll-like inability to fully support Th1 responses (16, 37, 38). We observed receptor function. Microbes Infect. 4:887. ␥ 13. Kaisho, T., O. Takeuchi, T. Kawai, K. Hoshino, and S. Akira. 2001. Endotoxin- a significant decrease in IFN- production by the CD8 T cells in induced maturation of MyD88-deficient dendritic cells. J. Immunol. 166:5688. vivo, which was previously reported for CD4 T cells in MyD88- 14. Kawai, T., O. Adachi, T. Ogawa, K. Takeda, and S. Akira. 1999. Unresponsive- deficient mice (16). We further show that expression of costimu- ness of MyD88-deficient mice to endotoxin. Immunity 11:115. Ϫ/Ϫ 15. Hacker, H., R. M. Vabulas, O. Takeuchi, K. Hoshino, S. Akira, and H. Wagner. latory molecules by DC from B6 or MyD88 is similar upon 2000. Immune cell activation by bacterial CpG-DNA through myeloid differen- treatment with hsp65-P1, in concordance with previous studies tiation marker 88 and tumor necrosis factor receptor-associated factor (TRAF)6. (15), thereby confirming that MyD88 is not required for this aspect J. Exp. Med. 192:595. 16. Schnare, M., G. M. Barton, A. C. Holt, K. Takeda, S. Akira, and R. Medzhitov. of DC activation. Therefore, it is not clear at present whether the 2001. Toll-like receptors control activation of adaptive immune responses. Nat. reduction of CTL activation in vivo is attributable to the display of Immun. 2:947. b 17. Lennon-Dumenil, A. M., A. H. Bakker, R. Maehr, E. Fiebiger, H. S. Overkleeft, fewer K /SIYRYYGL complexes at the DC surface, or whether M. Rosemblatt, H. L. Ploegh, and C. Lagaudriere-Gesbert. 2002. Analysis of other factors required for optimal CD8 T cell activation, such as protease activity in live -presenting cells shows regulation of the phago- decreased production of DC cytokines, are responsible. somal proteolytic contents during dendritic cell activation. J. Exp. Med. 196:529. 18. Trombetta, E. S., M. Ebersold, W. Garrett, M. Pypaert, and I. Mellman. 2003. Downloaded from For the activation of naive T cells, multiple interactions between Activation of lysosomal function during dendritic cell maturation. Science peptide/MHC and TCR are required (39–42). Naive CD8 T cells 299:1400. that recognize peptides presented by class I MHC molecules may 19. Delamarre, L., H. Holcombe, and I. Mellman. 2003. Presentation of exogenous on major histocompatibility complex (MHC) class I and MHC class II require further costimulatory signals for their activation (43–46). molecules is differentially regulated during dendritic cell maturation. J. Exp. Med. Once properly primed, such CD8 T cells need to recognize only 198:111. 1–10 peptides/class I MHC combinations to become functional 20. Sykulev, Y., Y. Vugmeyster, A. Brunmark, H. L. Ploegh, and H. N. Eisen. 1998. Peptide antagonism and T cell receptor interactions with peptide-MHC com-

CTLs, without the need for additional accessory signals (47–51). plexes. Immunity 9:475. http://www.jimmunol.org/ In the current study, all in vivo parameters were analyzed 3 days 21. Chu, N. R., H. B. Wu, T. Wu, L. J. Boux, M. I. Siegel, and L. A. Mizzen. 2000. Immunotherapy of a human papillomavirus (HPV) type 16 E7-expressing tumour after immunization. At this time point after immunization, the ex- by administration of fusion protein comprising Mycobacterium bovis bacille pression of SIYRYYGL-loaded class I molecules on the DC sur- Calmette-Guerin (BCG) hsp65 and HPV16 E7. Clin. Exp. Immunol. 121:216. face may be affected by the absence of MyD88. Our data under- 22. Adachi, O., T. Kawai, K. Takeda, M. Matsumoto, H. Tsutsui, M. Sakagami, K. Nakanishi, and S. Akira. 1998. Targeted disruption of the MyD88 gene results score this possibility, as the 2C CD8 T cells are activated in normal in loss of IL-1- and IL-18-mediated function. Immunity 9:143. fashion when exposed to SIYRYYGL/class I MHC on MyD88- 23. Kranz, D. M., S. Tonegawa, and H. N. Eisen. 1984. Attachment of an anti- deficient APCs in vivo, as inferred from their unchanged prolifer- receptor to non-target cells renders them susceptible to lysis by a clone of cytotoxic T lymphocytes. Proc. Natl. Acad. Sci. USA 81:7922. ative capacity and normal display of the CD44 activation marker. 24. Inaba, K., M. Inaba, N. Romani, H. Aya, M. Deguchi, S. Ikehara, S. Muramatsu, However, the MyD88 defect manifests itself in the inability of and R. M. Steinman. 1992. Generation of large numbers of dendritic cells from by guest on September 24, 2021 CD8 T cells to differentiate into fully activated CTL, as shown by mouse bone marrow cultures supplemented with granulocyte/macrophage colo- ny-stimulating factor. J. Exp. Med. 176:1693. the decrease in IFN-␥ production and strongly reduced CTL killing 25. Sha, W. C., C. A. Nelson, R. D. Newberry, J. K. Pullen, L. R. Pease, J. H. Russell, ability. The ability of DCs to fully engage and stimulate T cells is and D. Y. Loh. 1990. Positive selection of transgenic receptor-bearing thymo- cytes by Kb antigen is altered by Kb mutations that involve peptide binding. thus controlled at multiple levels, with an important role for Proc. Natl. Acad. Sci. USA 87:6186. MyD88 in cross-presentation. 26. Melief, C. J. 2003. Mini-review: regulation of cytotoxic T responses by dendritic cells: peaceful coexistence of cross-priming and direct priming? Eur. J. Immunol. 33:2645. Acknowledgments 27. Boes, M., N. Bertho, J. Cerny, M. Op Den Brouw, T. Kirchhausen, and We thank Evelyn Wiebe for determination of endotoxin content in H. Ploegh. 2003. T cells induce extended class II MHC compartments in dendritic hsp65-P1 and Carol McKinley for skillful technical assistance. We also cells in a Toll-like receptor-dependent manner. J. Immunol. 171:4081. thank Dr. Herman Eisen for critical reading of the manuscript and are 28. Poltorak, A., X. He, I. Smirnova, M. Y. Liu, C. Van Huffel, X. Du, D. Birdwell, E. Alejos, M. Silva, C. Galanos, et al. 1998. Defective LPS signaling in C3H/HeJ grateful to him for allowing a part of the work to be undertaken in his lab. and C57BL/10ScCr mice: mutations in TLR4 gene. Science 282:2085. 29. Kane, L. P., J. Lin, and A. Weiss. 2002. It’s all Rel-ative: NF-␬B and CD28 References costimulation of T-cell activation. Trends Immunol. 23:413. 1. Sigal, L. J., and K. L. Rock. 2000. Bone marrow-derived antigen-presenting cells 30. Vogt, A. B., S. Spindeldreher, and H. Kropshofer. 2002. Clustering of MHC- are required for the generation of cytotoxic T lymphocyte responses to peptide complexes prior to their engagement in the immunological synapse: lipid and use transporter associated with antigen presentation (TAP)-dependent and raft and tetraspan microdomains. Immunol. Rev. 189:136. -independent pathways of antigen presentation. J. Exp. Med. 192:1143. 31. Tulp, A., D. Verwoerd, B. Dobberstein, H. L. Ploegh, and J. Pieters. 1994. Iso- 2. Kurts, C. 2000. Cross-presentation: inducing CD8 T cell immunity and tolerance. lation and characterization of the intracellular MHC class II compartment. Nature J. Mol. Med. 78:326. 369:120. 3. Heath, W. R., and F. R. Carbone. 2001. Cross-presentation, dendritic cells, tol- 32. Amigorena, S., P. Webster, J. Drake, J. Newcomb, P. Cresswell, and I. Mellman. erance and immunity. Annu. Rev. Immunol. 19:47. 1995. Invariant chain cleavage and peptide loading in major histocompatibility 4. Suzue, K., X. Zhou, H. N. Eisen, and R. A. Young. 1997. Heat shock fusion complex class II vesicles. J. Exp. Med. 181:1729. proteins as vehicles for antigen delivery into the major histocompatibility com- 33. Schnare, M., G. M. Barton, A. C. Holt, K. Takeda, S. Akira, and R. Medzhitov. plex class I presentation pathway. Proc. Natl. Acad. Sci. USA 94:13146. 2001. Toll-like receptors control activation of adaptive immune responses. Nat. 5. Chu, N. R., H. B. Wu, T. Wu, L. J. Boux, M. I. Siegel, and L. A. Mizzen. 2000. Immun. 2:947. Immunotherapy of a human papillomavirus (HPV) type 16 E7-expressing tumor 34. Lord, K. A., B. Hoffman-Liebermann, and D. A. Liebermann. 1990. Nucleotide by administration of fusion protein comprising Mycobacterium bovis bacille sequence and expression of a cDNA encoding MyD88, a novel myeloid differ- Calmette-Guerin (BCG) hsp65 and HPV16 E7. Clin. Exp. Immunol. 121:216. entiation primary response gene induced by IL6. Oncogene 5:1095. 6. Cho, B. K., D. Palliser, E. Guillen, J. Wisniewski, R. A. Young, J. Chen, and 35. Yamamoto, M., S. Sato, H. Hemmi, K. Hoshino, T. Kaisho, H. Sanjo, H. N. Eisen. 2000. A proposed mechanism for the induction of cytotoxic T lym- O. Takeuchi, M. Sugiyama, M. Okabe, K. Takeda, and S. Akira. 2003. Role of phocyte production by heat shock fusion proteins. Immunity 12:263. adaptor TRIF in the MyD88-independent Toll-like receptor signaling pathway. 7. Bennett, S. R., F. R. Carbone, F. Karamalis, R. A. Flavell, J. F. Miller, and Science 301:640. W. R. Heath. 1998. Help for cytotoxic-T-cell responses is mediated by CD40 36. Hoebe, K., X. Du, P. Georgel, E. Janssen, K. Tabeta, S. O. Kim, J. Goode, P. Lin, signalling. Nature 393:478. N. Mann, S. Mudd, et al. 2003. Identification of Lps2 as a key transducer of 8. Ridge, J. P., F. Di Rosa, and P. Matzinger. 1998. A conditioned dendritic cell can MyD88-independent TIR signalling. Nature 424:743. be a temporal bridge between a CD4ϩ T-helper and a T-killer cell. Nature 37. Jankovic, D., M. C. Kullberg, S. Hieny, P. Caspar, C. M. Collazo, and A. Sher. 393:474. 2002. In the absence of IL-12, CD4ϩ T cell responses to intracellular pathogens The Journal of Immunology 3421

fail to default to a Th2 pattern and are host protective in an IL-10Ϫ/Ϫ setting. 45. McAdam, A. J., E. A. Farkash, B. E. Gewurz, and A. H. Sharpe. 2000. B7 Immunity 16:429. costimulation is critical for antibody class switching and CD8ϩ cytotoxic T- 38. Muraille, E., C. De Trez, M. Brait, P. De Baetselier, O. Leo, and Y. Carlier. 2003. lymphocyte generation in the host response to vesicular stomatitis . J. Virol. Genetically resistant mice lacking MyD88-adapter protein display a high suscep- 74:203. tibility to Leishmania major infection associated with a polarized Th2 response. 46. Lumsden, J. M., J. M. Roberts, N. L. Harris, R. J. Peach, and F. Ronchese. 2000. J. Immunol. 170:4237. Differential requirement for CD80 and CD80/CD86-dependent costimulation in 39. Valitutti, S., S. Muller, M. Cella, E. Padovan, and A. Lanzavecchia. 1995. Serial the lung immune response to an influenza virus infection. J. Immunol. 164:79. triggering of many T-cell receptors by a few peptide-MHC complexes. Nature 47. Brower, R. C., R. England, T. Takeshita, S. Kozlowski, D. H. Margulies, 375:148. J. A. Berzofsky, and C. Delisi. 1994. Minimal requirements for peptide mediated 40. Underhill, D. M., M. Bassetti, A. Rudensky, and A. Aderem. 1999. Dynamic activation of CD8ϩ CTL. Mol. Immunol. 31:1285. interactions of macrophages with T cells during antigen presentation. J. Exp. 48. Kageyama, S., T. J. Tsomides, Y. Sykulev, and H. N. Eisen. 1995. Variations in Med. 190:1909. the number of peptide-MHC class I complexes required to activate cytotoxic T 41. Gunzer, M., A. Schafer, S. Borgmann, S. Grabbe, K. S. Zanker, E. B. Brocker, cell responses. J. Immunol. 154:567. E. Kampgen, and P. Friedl. 2000. Antigen presentation in extracellular matrix: interactions of T cells with dendritic cells are dynamic, short lived, and sequen- 49. Wang, W., P. H. Gulden, R. A. Pierce, J. Shabanowitz, S. T. Man, D. F. Hunt, tial. Immunity 13:323. V. H. Engelhard, and J. A. Shabanowitz. 1997. A naturally processed peptide presented by HLA-A*0201 is expressed at low abundance and recognized by an 42. Irvine, D. J., M. A. Purbhoo, M. Krogsgaard, and M. M. Davis. 2002. Direct ϩ observation of ligand recognition by T cells. Nature 419:845. alloreactive CD8 cytotoxic T cell with apparent high affinity. J. Immunol. 43. Shahinian, A., K. Pfeffer, K. P. Lee, T. M. Kundig, K. Kishihara, A. Wakeham, 158:5797. K. Kawai, P. S. Ohashi, C. B. Thompson, and T. W. Mak. 1993. Differential T 50. Sykulev, Y., M. Joo, I. Vturina, T. J. Tsomides, and H. N. Eisen. 1996. Evidence cell costimulatory requirements in CD28-deficient mice. Science 261:609. that a single peptide-MHC complex on a target cell can elicit a cytolytic T cell 44. Kundig, T. M., A. Shahinian, K. Kawai, H. W. Mittrucker, E. Sebzda, response. Immunity 4:565. M. F. Bachmann, T. W. Mak, and P. S. Ohashi. 1996. Duration of TCR stimu- 51. Viola, A., and A. Lanzavecchia. 1996. T cell activation determined by T cell lation determines costimulatory requirement of T cells. Immunity 5:41. receptor number and tunable thresholds. Science 273:104. Downloaded from http://www.jimmunol.org/ by guest on September 24, 2021