Murine Or Human XCR1 Cross-Presenting Dendritic Cells In

Induction of Potent CD8 T Cell Cytotoxicity by Specific Targeting of Antigen to Cross-Presenting Dendritic Cells In Vivo via Murine or Human XCR1 This information is current as of October 1, 2021. Evelyn Hartung, Martina Becker, Annabell Bachem, Nele Reeg, Anika Jäkel, Andreas Hutloff, Harald Weber, Christoph Weise, Claudia Giesecke, Volker Henn, Stephanie Gurka, Konstantinos Anastassiadis, Hans W. Mages and Richard A. Kroczek Downloaded from J Immunol 2015; 194:1069-1079; Prepublished online 17 December 2014; doi: 10.4049/jimmunol.1401903 http://www.jimmunol.org/content/194/3/1069 http://www.jimmunol.org/

References This article cites 35 articles, 14 of which you can access for free at: http://www.jimmunol.org/content/194/3/1069.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision by guest on October 1, 2021

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

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

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

Induction of Potent CD8 T Cell Cytotoxicity by Speciﬁc Targeting of Antigen to Cross-Presenting Dendritic Cells In Vivo via Murine or Human XCR1

Evelyn Hartung,*,1 Martina Becker,*,1 Annabell Bachem,* Nele Reeg,* Anika Ja¨kel,* Andreas Hutloff,* Harald Weber,* Christoph Weise,† Claudia Giesecke,‡ Volker Henn,* Stephanie Gurka,* Konstantinos Anastassiadis,x Hans W. Mages,* and Richard A. Kroczek*

Current subunit vaccines are incapable of inducing Ag-speciﬁc CD8+ T cell cytotoxicity needed for the defense of certain infections and for therapy of neoplastic diseases. In experimental vaccines, cytotoxic responses can be elicited by targeting of Ag into cross-

presenting dendritic cells (DC), but almost all available systems use target molecules also expressed on other cells and thus lack the Downloaded from desired specificity. In the present work, we induced CD8+ T cell cytotoxicity by targeting of Ag to XCR1, a chemokine receptor exclusively expressed on murine and human cross-presenting DC. Targeting of Ag with a mAb or the chemokine ligand XCL1 was highly specific, as determined with XCR1-deficient mice. When applied together with an adjuvant, both vector systems induced a potent cytotoxic response preventing the outgrowth of an inoculated aggressive tumor. By generating a transgenic mouse only expressing the human XCR1 on its cross-presenting DC, we could demonstrate that targeting of Ag using human XCL1 as vector is fully effective in vivo. The specificity and efficiency of XCR1-mediated Ag targeting to cross-presenting DC, combined with its http://www.jimmunol.org/ lack of adverse effects, make this system a prime candidate for the development of therapeutic cytotoxic vaccines in humans. The Journal of Immunology, 2015, 194: 1069–1079.

odern subunit vaccines stimulate the B cell immune Apart from prophylactic vaccination, a therapeutic vaccination system and induce neutralizing Abs. However, neu- employing the cytotoxic potential of the immune system is also needed M tralizing Abs cannot be effective against pathogens that due to the limitations of current therapies against many hematological “hide” intracellularly (e.g., Plasmodium malariae), or that per- and solid tumors. In particular, an immune therapy directed at re- manently change the antigenic nature of their surface structures moving residual cancerous tissue or metastases after surgical or by guest on October 1, 2021 (“antigenic drift”), as is the case with seasonal influenza or HIV. chemical ablation of the tumor would be highly desirable. Therefore, there is a need for vaccines utilizing the second pow- The repertoire of CD8+ T cells in a naive organism contains erful arm of the adaptive immune system, the cytotoxic defense only relatively few cells of a given Ag specificity, estimated to be + (1). In this type of immunity, CD8 T cells specifically recognize in the order of several hundred cells (2). To induce potent cyto- peptide fragments derived from (often conserved) pathogen pro- toxicity starting from this low number of Ag-specific T cells, teins that are presented on the surface of infected cells in the a very effective Ag presentation to CD8+ T cells in an inflam- + context of MHC I. As a result, the CD8 T cells become activated, matory context is required. Only dendritic cells (DC) as profes- acquire cytotoxic capacity, and eliminate these infected cells in sional APCs can prime naive CD8+ T cells to become effectors. It a highly specific fashion. is for this reason that elegant strategies to directly introduce Ag into DC in vivo have been developed in the past, originally in the laboratories of Steinman and Nussenzweig (3, 4), and later by

*Molecular Immunology, Robert Koch-Institute, 13353 Berlin, Germany; †Institute others (5–8). All of these approaches to effectively induce cyto- + of Chemistry and Biochemistry, Free University, 14195 Berlin, Germany; ‡Depart- toxic CD8 T cells were based on Ag coupled to Abs directed to ment of Medicine, Rheumatology, and Clinical Immunology, Charite´ University x surface molecules on DC. Medicine, 10117 Berlin, Germany; and Biotechnology Center, Technical Univer- sity, 01307 Dresden, Germany Conventional DC in the mouse are currently being subdivided 1E.H. and M.B. contributed equally to this work. into two populations. DC dependent in their development on the Received for publication July 25, 2014. Accepted for publication November 17, transcription factor Batf3 are highly capable of presenting ex- + 2014. ogenous Ag to naive CD8 T cells (cross-presentation) (9–12), + This work was supported by grants of the Wilhelm Sander Foundation, the Fritz and Batf3-independent DC excel in Ag presentation to CD4 Thyssen Foundation, the Validation of Innovation program of the Bundesministerium T cells (10, 13). To optimally induce cytotoxic immunity, specific fur€ Bildung und Forschung, and in part by Deutsche Forschungsgemeinschaft grants to R.A.K. (Kr 827/16-1 and Kr 827/18-1). targeting of Ag into the Batf3-dependent DC is therefore highly Address correspondence and reprint requests to Prof. Richard A. Kroczek, Molecular desirable. Most of the original targeting studies could not meet Immunology, Robert Koch-Institute, Nordufer 20, 13353 Berlin, Germany. E-mail this goal, because at that time surface molecules restricted in their address: [email protected] expression to this cross-presenting DC subset were not yet Abbreviations used in this article: DC, dendritic cell; Dig, digoxigenin; LN, lymph known. In consequence, most of the work has been done on node; RT, room temperature. targeting Ag via CD205, a C-type lectin receptor present on Copyright Ó 2015 by The American Association of Immunologists, Inc. 0022-1767/15/$25.00 cross-presenting DC, but also expressed on a variety of cell types www.jimmunol.org/cgi/doi/10.4049/jimmunol.1401903 1070 SPECIFIC Ag TARGETING TO CROSS-PRESENTING DC in mice (14, 15) and humans (16). Although the desired targeting construct for XCL1-StrepTag was designed to contain murine XCL1 and specificity could not yet be achieved, these studies were pivotal in StrepTag (SAWSHPQFEKGGGSGGGSGGGS WSHPQFEK), separated that they demonstrated that effective CD8+ T cell cytotoxicity by a glycine-serine linker (GS). The DNA for XCL1-StrepTag was chemically synthesized (Geneart/Life Technologies) and cloned into could be induced through direct delivery of Ag into DC in vivo. pRmHa-3. XCL1-OVA was generated by cloning murine XCL1 and full- In the past few years, we and others recognized that the chemokine length chicken OVA, separated by a glycine-serine linker (GS), into receptor XCR1 is exclusively expressed on the Batf3-dependent, cross- pRmHa-3. For generating recombinant MARX10-OVA Ab, we first de- presenting DC and not elsewhere in the body (17–19). This highly termined the DNA sequences coding for the complete k L chain and the V region of the H chain of mAb MARX10. The DNA for the k L chain was restricted expression of XCR1 in the mouse is mirrored in the human chemically synthesized (Eurofins Genomics) in frame with a DNA se- system, where XCR1 can only be found on DC homologous to the quence coding for a signal peptide and cloned into the mammalian ex- mouse cross-presenting subset (18, 20, 21). XCR1 thus appears to be pression vector pTT5 (NRC-Biotechnology Research Institute). The H an ideal target to induce cytotoxic immunity in the human. chain of MARX10-OVAwas generated as follows: as a backbone, we used In the present work, we targeted Ag to XCR1 in the mouse using the DNA sequence of DEC-OVA (provided by M.C. Nussenzweig, Rockefeller University, NY) that was cloned via EcoRI/NotI into pTT5. either a mAb or the specific murine chemokine ligand XCL1. Both The V region of the MARX10 H chain was chemically synthesized in vectors delivered Ag into cross-presenting DC in a highly specific frame with a sequence for a signal peptide, including an EcoRI restriction fashion and induced, when applied with LPS as adjuvant, potent site at its 59 end and a short sequence of the CH1 domain of DEC-OVA, cytotoxic immunity. Moreover, we generated a transgenic mouse including a unique BstEII restriction site at its 39 end (Eurofins Genomics). The synthesized DNA fragment was then used to replace the V region of that orthotopically expresses only the human XCR1 receptor on the H chain of the DEC-OVA Ab with that of MARX10 via the restriction cross-presenting DC. In this model, we could demonstrate efficient sites EcoRI/BstEII. huXCL1-StrepTag was designed to contain DNA induction of cytotoxic immunity with human XCL1 or XCL2 as for human XCL1 and StrepTag (SAWSHPQFEKGGGSGGGSGGSA- vectors. The successful targeting of Ag into cross-presenting DC WSHPQFEK), separated by a glycine (G). The DNA for huXCL1-StrepTag Downloaded from via the human XCR1 under physiological conditions is a major step was chemically synthesized (Eurofins Genomics) and cloned into pRmHa-3. For generating huXCL2-StrepTag, human XCL2 (differing from huXCL1 at forward in establishing this system for induction of therapeutic positions 28(D→H) and 29(K→R)) was chemically synthesized by Eurofins CD8+ T cell cytotoxicity against neoplastic diseases in the human. Genomics and used to replace huXCL1 in the huXCL1-StrepTag plasmid. huXCL1-OVA was generated by fusing human XCL1 via a glycine-serine Materials and Methods linker (GS) to full-length chicken OVA, followed by the TEV-cleavage site

Mice ENLYFQG (separated by a GS linker) and the StrepTag SAWSHPQ- http://www.jimmunol.org/ FEKGGGSGGGSGGSAWSHPQFEK (separated by a GGSGGSG linker). C57BL/6 mice (8–10 wk old) were used for experiments and cell isolation, The whole construct was cloned into the pRmHa-3 plasmid. huXCL2 was unless indicated otherwise. B6.XCR1-lacZ mice (The Jackson Laboratory) constructed in an analogous manner. is a strain in which the XCR1 gene has been replaced by the b-Gal reporter gene; these mice were fully backcrossed (.103) onto the C57BL/6 Expression and purification of recombinant proteins background. OT-I and OT-II TCR-transgenic mice were crossed onto the XCL1, XCL1-StrepTag, huXCL1-StrepTag, huXCL2-StrepTag, XCL1- B6.PL background to allow identification of CD8+ and CD4+ T cells using OVA, huXCL1-OVA, or huXCL2-OVA encoding plasmids were electro- the CD90.1 marker. All mice were bred under specific pathogen-free porated together with the plasmid phshs.PURO into drosophila Schneider conditions in the animal facility of the Federal Institute for Risk Assess- SL-3 cells (24) using a Bio-Rad Gene Pulser (450 V and 500 mF). The ment (Berlin, Germany). Experiments were performed according to state phshs.PURO plasmid (provided by M. McKeown, Salk Institute) allows guidelines and approved by the local animal welfare committee. by guest on October 1, 2021 selection of positive transfectants by puromycin. Clones from limiting Generation of huXCR1 transgenic mice dilution cultures of transfected SL-3 cells were analyzed for high protein production using either XCL1- or StrepTag-specific ELISA and expanded A BAC containing the human XCR1 gene was obtained from Children’s in serum-free Insect-XPRESS medium (Lonza) on a shaker platform (100 Hospital Oakland Research Institute (clone ID: RP11-527M10). BAC rpm) in normal air at 27˚C. XCL1 protein was purified from supernatants DNA was prepared from bacterial culture grown at 30˚C using the by heparin Sepharose affinity chromatography, followed by reversed-phase Nucleobond BAC 100 kit (Macherey-Nagel), digested with MluI/AscI for HPLC; the identity of the protein was verified by mass spectrometry. removing the bacterial backbone, purified by Sepharose CL-4B chroma- XCL1-StrepTag protein was purified from the supernatant using HiTrap tography (22), and injected (1–4 ng/ml) into the pronucleus of C57BL/6J Heparin HP columns (GE Healthcare). XCL1/2-OVA protein was purified fertilized oocytes. Offspring was genotyped using the primer pair 59- by heparin Sepharose affinity chromatography, followed by anion ex- GCACGGTCAAGCTCATCTTCGCC-39 and 59-ATGCTCCTTCCAGGC- change chromatography on a Mono Q 5/50 GL column (GE Healthcare). CCGCT-39. Five transgenic founders (which expressed both wild-type and huXCL1-StrepTag, huXCL2-StrepTag, and huXCL1-OVA were purified human XCR1) were crossed with homozygous B6.XCR1-lacZ mice using StrepTrap HP columns (GE Healthcare). pTT5 plasmids encoding (deficient for XCR1) to obtain mice only expressing human XCR1. The the H and L chains of recombinant MARX10-OVA were transiently founder line with the highest huXCR1 orthotopic expression was then cotransfected into 293-6E cells (NRC-Biotechnology Research Institute) chosen for further experiments (B6.huXCR1tg). with linear polyethylenimine. Cells were grown in serum-free FreeStyle Cells and cell isolation F17 expression medium (Life Technologies) supplemented with 0.1% Pluronic F68, 4 mM L-glutamine, and 25 mg/ml G418 on a shaker platform Murine splenocytes were obtained by mashing spleens through 70-mm cell (125 rpm) in 5% CO2 at 37˚C. mAb MARX10-OVA was purified from the sieves into PBS, followed by erythrocyte lysis with ACK buffer (155 mM supernatant using HiTrap Protein G HP columns (GE Healthcare). LPS NH4Cl, 10 mM KHCO3, 0.1 mM EDTA). Where indicated and with lymph content of XCL1 (0.26 EU/mg), XCL1-OVA (12 EU/mg used for che- nodes (LN), DC were enriched by digesting pieces of lymphoid tissue with motaxis and test of DC activation in vivo; .50 EU/mg used for in vivo collagenase D (500 mg/ml) and DNase I (20 mg/ml; both Roche) in RPMI proliferation and CTL induction), and MARX10-OVA (,0.6 EU/mg) was 1640 containing 2% FCS (low endotoxin; Biochrom) for 20 min at 37˚C; determined using the Endosafe-PTS test (Charles River). EDTA (10 mM) was added for additional 5 min, and cells were filtered through a 70-mm nylon sieve (BD Falcon). Human PBMC and splenocytes were Antibodies prepared by standard Biocoll (Biochrom) density gradient centrifugation of buffy coats or mashed spleens, respectively. Human splenocytes were from Hybridomas producing mAb recognizing CD4 (clone YTS 191.1), CD8 splenectomies in patients suffering from spherocytosis or massive cysts; their (53-6.72), CD11c (N418), CD16/32 (2.4G2), CD45R/B220 (RA3-6B2), use in this study was approved by the local ethics committee. CD80 (16-10A1), CD86 (GL1), IFN-g (AN18.17.24), Ly6G/C (RB6-8C5), SIINFEKL-H-2Kb (25-D1.16), and MHC II (M5/114.15.2) were ob- Cloning of XCL1, XCL1-StrepTag, XCL1-OVA, MARX10-OVA, tained from American Type Culture Collection, and CD90.1 (OX-7) from huXCL1-StrepTag, huXCL1-OVA, huXCL2-StrepTag, and European Cell Culture Collection. mAb to CD103 (M290) was from BD huXCL2-OVA Biosciences, and to CD69 (H1.2F3) and PD-L2 (TY25) from eBioscience. Anti-CD3 (KT3) was generously provided by H. Savelkoul (University The DNA coding for full-length murine XCL1 was cloned into the dro- of Wageningen, Wageningen, the Netherlands), anti-CD25 (2E4) by sophila expression vector pRmHa-3 (23) by standard procedures. The E. Shevach (National Institute of Allergy and Infectious Diseases, National The Journal of Immunology 1071

Institutes of Health, Bethesda, MD), and anti-CD40 (FGK) by T. Rolink for 4.5 h, with 5 mg/ml brefeldin A (all from Sigma-Aldrich) present for (University of Basel, Basel, Switzerland). Generation of anti-XCR1 mAb the last 3.5 h. Cells were then stained for surface Ags, labeled with 1.34 MARX10 (19) and anti-ICOS mAb (MIC-280 (25)) has been described mM Pacific Orange to mark dead cells for later gating, and fixed with 2% before. The nonagonistic mAb MARX10 (mouse IgG2b, in the recombi- paraformaldehyde. Intracellular staining was performed at RT in a solution nant version IgG1) does not block the binding of XCL1 to XCR1. For of 0.5% saponin (Sigma-Aldrich), 100 mg/ml 2.4G2, and 50 mg/ml purified staining of human PBMC, the following Abs were used: CD19 (clone rat Ig. LPS was removed from OVA using EndoTrap red (Hyglos), resulting BU12) (26), CD3 (OKT3), CD14 (63D3), and HLA-DR (L243) from in ,0.5 EU endotoxin/mg protein as determined by the Endosafe-PTS assay American Type Culture Collection; CD11c (BU15), CD16 (3G8), and (Charles River). CD56 (HCD56) from BioLegend; and CD141 (AD5-14H12) from Milte- nyi Biotec. mAb D3-13F1 (IgG2b, specific for nitrophenol, a gift of Determination of OVA-specific Abs in serum K. Rajewsky) was used as the isotype control for the natural mAb MARX10. C57BL/6 or homozygous B6.XCR1-lacZ mice were immunized i.v. with Flow cytometry OVA (2 mg), MARX10-OVA (0.16 mg), or XCL1-OVA (1 mg), together with 3 mg LPS; controls received only LPS or PBS. After 14 d, OVA- Abs were titrated for optimal signal-to-noise ratio. Mouse DC were specific Ab in the serum were determined by ELISA. In short, flat-bottom + + 2 identified as MHC II CD11c Lin (CD3, B220, Ly6G/C), and human DC MaxiSorp 96-well immunoplates (Nunc) were coated with 100 mg/ml OVA + + 2 as MHC II CD11c Lin (CD3, CD14, CD16, CD19, CD56). For staining in PBS overnight at 4˚C, blocked with 1% BSA in PBS for 1 h at RT, and of XCR1, mAb MARX10 was directly coupled to PE, and MARX10-OVA incubated with serum dilutions in PBS, 1% BSA for 1 h at 37˚C. OVA- was detected with goat anti-mouse IgG Cy5 (Dianova). XCL1-StrepTag, specific Ig isotypes were measured with goat anti-mouse IgM-, IgG1-, XCL1-OVA-StrepTag, huXCL1-StrepTag, and huXCL2-StrepTag were IgG2a-, and Ig2b-specific Abs coupled to HRP (Southern Biotech Anti- reacted with Strep-Tactin-PE (IBA) before being used for staining. To bodies), and 3,39,5,59-tetramethylbenzidine (Sigma-Aldrich) was used as block unspecific binding to Fc receptors, murine cells were preincubated substrate. with mAb 2.4G2 (20 mg/ml), and human PBMC with Endobulin (1 mg/ml;

Baxter Hyland-Immuno Division). Standard staining with mAb was in In vivo cytotoxicity assay Downloaded from PBS, 0.25% BSA, and 0.1% NaN3 for 20 min on ice. For digoxigenin (Dig)-coupled and biotinylated mAb, AF647-conjugated polyclonal sheep Animals were immunized with the indicated amounts of either MARX10- anti-Dig Abs (Roche) or streptavidin conjugated to PE, PE-Cy7, or allo- OVA, XCL1-OVA, huXCL1-OVA, or huXCL2-OVA i.v. Six days later, 5 3 phycocyanin (BioLegend) were used as secondary reagents. Data were 106 syngeneic splenocytes were pulsed with SIINFEKL peptide (Gen- acquired on a LSR II or Fortessa flow cytometers (both BD Biosciences), Script) and labeled with 10 mM CFSE (CFSEhigh) in vitro, and 5 3 106 and analyzed using FlowJo (Tree Star). splenocytes were left unpulsed and labeled with 1 mM CFSE (CFSElow). Both preparations were injected together i.v. into immunized and control + In vivo labeling of XCR1 DC animals, and the CFSE signal was determined by flow cytometry 18 h later. http://www.jimmunol.org/ Specific lysis was calculated using the following formula: specific lysis mAb MARX10 was coupled to Dig and injected i.v. (1.25 mg for flow high low high (%) = 100 2 ([CFSE immunized/CFSE immunized]/[CFSE cytometry or 10 mg for histology), and isotype mAb D3-13F1-Dig or PBS low control/CFSE control]) 3 100. were injected as controls; detection of Dig was with AF647-conjugated polyclonal sheep anti-Dig Ab (Roche). For histological analysis, cryostat Tumor model sections (12 mm) of splenic tissues obtained 16 h after injection of the respective mAb were fixed in acetone for 10 min at room temperature Six days after immunization, C57BL/6 mice were injected s.c. with 6 3 105 (RT); endogenous peroxidase was inhibited with a blocking solution (PBS, E.G7 cells (EL4 thymoma transfected with soluble OVA) (27) diluted in 1 mM NaN3, 10 mM glucose, 1 U/ml glucose oxidase) for 1 h at 37˚C; and 50 ml PBS. Because the percentage of E.G7 cells expressing SIINFEKL unspecific binding sites were saturated with casein solution (Vector Lab- on the surface decreases over time, the SIINFEKL+ E.G7 cells were flow oratories) for 1 h at RT. Sections were stained with FITC-conjugated mAb sorted 3 d before injection using mAb 25-D1.16. Tumor growth was by guest on October 1, 2021 RA3-6B2, washed, and incubated with anti-Dig Fab coupled to alkaline continuously monitored, and tumor size and weight were determined 14 d phosphatase and anti-fluorescein Fab coupled to HRP (both Roche). The after E.G7 cell injection. stainings were developed using the Blue Alkaline Phosphatase Substrate Kit and the ImmPACT DAB peroxidase substrate (both Vector Laborato- Statistical analysis ries) sequentially. Statistical analysis was performed using GraphPad Prism 5. Data are shown Chemotaxis assay as mean or mean 6 SEM. DC were obtained from digested murine splenic tissue (see cell isolation), and further enriched by centrifugation over a Nycoprep density gradient Results (1.073 g/ml), followed by positive magnetic sorting with anti-CD11c mAb MARX10 specifically binds to the murine chemokine re- 6 microbeads (Miltenyi Biotec). For chemotaxis assays, 1 3 10 DC (pu- ceptor XCR1 (19). To assess whether this mAb can bind to XCR1 ∼ rity 70%) were suspended in 100 ml chemotaxis medium (RPMI 1640, in vivo, mAb MARX10 was labeled with Dig and injected i.v. 1% BSA, 50 mM b-ME, 100 mg/ml penicillin/streptomycin) and placed into the upper chamber of a 24-well Transwell system (6.5 mm diameter, Lymphoid tissues were removed at various time points and ana- 5-mm pore polycarbonate membrane; Corning Costar). The lower chamber lyzed for the presence of the mAb on the surface of cells. mAb was filled with chemotaxis medium to which recombinant XCL1 or XCL1- MARX10 could be detected on splenic XCR1+ DC, but not on OVA was added. After incubation in 5% CO2 for 2 h at 37˚C, the number other splenic DC populations, as early as 10 min and as late as of migrated DC was determined by counting cells in the lower chamber using a flow cytometer. DC were identified by staining for XCR1, CD8, 1 wk after injection (Fig. 1A), the latter suggesting a long t1/2 of CD11c, and MHC II after gating out cells expressing B220 and CD3. The the mAb in the circulation. Only B cells (0.5%) showed some percentage of migrated cells was calculated by dividing the number of unspecific staining in the first 4 h, but were later negative (data not cells in the lower chamber by the number of input cells (number migrated shown). No signal was detected on DC in spleens of mice deficient 3 cells/number input cells 100). for XCR1 (Fig. 1B). In peripheral LN, MARX10 was found on the In vivo proliferation and in vitro activation of OT-I and OT-II surface of most resident XCR1+ DC (MHC IIint), but only on T cells a small population of migratory XCR1+ DC (MHC IIhigh)14h after injection (Fig. 1C), indicating that the mAb had good access Recipient mice were adoptively transferred with splenocytes containing 2 3 106 OT-I (CD8+) resting T cells (negative for CD25, CD69, and ICOS), to lymphoid tissues within the 14-h experimental period, but or with 1 3 106 naive transgenic OT-II (CD4+) T cells enriched from incomplete access to DC residing in peripheral organs. When splenocytes by positive sorting with CD62L beads (Miltenyi Biotec). For splenic tissue sections were analyzed for the presence of mAb proliferation analysis, OT-I and OT-II cells were labeled with 5 mM CFSE MARX10, the pattern obtained (Fig. 1D) was indicative of a (Invitrogen) before transfer and analyzed 48 h after immunization using + the CFSE dilution assay. For measuring IFN-g production, splenocytes highly specific binding to XCR1 DC (17, 19). Similar studies from immunized mice were isolated 72 h after immunization and stimu- with the chemokine ligand XCL1 did not give a signal (our un- lated with 0.02 mg/ml PMA and 1 mg/ml ionomycin in complete medium published data). 1072 SPECIFIC Ag TARGETING TO CROSS-PRESENTING DC Downloaded from http://www.jimmunol.org/ by guest on October 1, 2021

FIGURE 1. Binding characteristics of mAb MARX10 in vivo. MARX10-Dig (1.25 mg) was injected i.v. into C57BL/6 (A, C, D) or B6.XCR1-lacZ+/+ (B) mice, and controls received a Dig-coupled isotype mAb (A, C) or PBS (D). At the indicated time points, animals were sacrificed and the presence of mAb MARX10 bound to splenocytes was visualized using a fluorophore-coupled anti-Dig reagent in flow cytometry (A–C). To correlate the anti-Dig signal with expression of XCR1, the splenocytes were costained with mAb MARX10-PE in vitro (A–C). Shown are the flow cytometry profiles of splenic (A and B) and skin-draining LN (C) DC. For histological analysis, MARX10-Dig (10 mg) was injected i.v. and splenic tissue was analyzed for bound Ab (blue) 16 h after injection; counterstaining (brown) was with B cell–specific mAb RA3-6B. Scale bar, 100 mm(D). The kinetics experiment was performed once, and all other experiments twice.

Recombinant MARX10-OVA and XCL1-OVA specifically bind MARX10-OVA, XCL1-OVA, and XCL1 do not change the to XCR1 activation status of XCR1+ DC in vivo To target Ag to cross-presenting DC via XCR1 in vivo, we gen- To test whether recombinant XCL1-OVA is still capable of in- erated two delivery systems based on mAb MARX10 and XCL1, ducing chemotaxis, Transwell experiments were performed with the specific chemokine ligand for XCR1. In a first step, the Ag- splenic DC. Various amounts of XCL1-OVA induced chemotaxis binding regions of mAb MARX10 were grafted onto the back- of up to 30% of XCR1+ DC, whereas XCR12 DC did not migrate; bone of mAb DEC-205, which has previously been point mutated to the functional effects of XCL1-OVA were thus comparable to the minimize binding to Fc receptors (3). The chimeric mAb MARX10 parent chemokine XCL1 (Fig. 3A). The targeting vector MARX10- was then modified to accommodate OVA as a C-terminal fusion OVA did not induced chemotaxis in analogous experiments (our protein (Fig. 2A). Recombinant versions of XCL1 and XCL1- unpublished data). OVA (Fig. 2A, all drawn to scale) were generated in a separate Because XCL1 and XCL1-OVA function as chemoattractants for approach. When tested in vitro, both MARX10-OVA and XCL1- XCR1+ DC, we assessed whether their use in vivo changes the OVA specifically bound to XCR1+ DC, but not to DC from XCR1- activation state of XCR1+ DC. MARX10-OVA was also included deficient mice (Fig. 2B), or any other cells (data not shown). in this analysis, although this nonblocking mAb does not have The Journal of Immunology 1073

FIGURE 2. MARX10-OVA and XCL1-OVA specifically bind to XCR1. (A) The V region of the H chain and the V and C regions of the L chain of mAb MARX10 (blue) were grafted onto the backbone of mAb DEC-205 (white), and the entire sequence of OVA (OVA, red) fused to the C terminus of this construct (mAb MARX10-OVA). OVA was also fused to the C terminus of murine XCL1. (B) Splenocytes from C57BL/6 mice (red histograms) or from mice deficient for XCR1 (homozygous B6.XCR1-lacZ mice, filled gray histograms) were stained with MARX10-OVA or XCL1-OVA, and for comparison with the parent molecules MARX10 or XCL1. Shown are the staining profiles after gating on DC defined as CD11c+ MHC II+ Lin2 cells. Representative

experiment of .10 (MARX10, XCL1) and two (MARX10-OVA, XCL1-OVA). Downloaded from agonist activity. MARX10-OVA and XCL1-OVA were injected i.v. at elicits a short Ca2+ burst after binding to its receptor (17), did not amounts later used for Ag targeting in vivo, XCL1 at higher amounts. result in a sustained activation of XCR1+ DC in vivo (Fig. 3).

Spleens of treated mice were removed 14 h later and analyzed for + expression levels of CD80, CD86, CD40, and PD-L2, the most MARX10-OVA and XCL1-OVA induce proliferation of CD4 T cells, their differentiation to Th1 cells, and the generation of sensitive markers of DC activation. None of the targeting reagents http://www.jimmunol.org/ injected changed the expression levels of the activation markers, isotype-switched Ag-speciﬁc Abs in vivo which remained at the steady state levels observed after injection of XCR1+ DC not only cross-present, but also classically present Ags PBS (Fig. 3). Even injection of substantial amounts of XCL1, which in the context of MHC II. To examine whether Ag targeted via by guest on October 1, 2021

FIGURE 3. Effects of the targeting systems MARX10-OVA and XCL1-OVA on XCR1+ DC. (A) XCL1-OVA, similar to its parent molecule XCL1, selectively induced chemotaxis of XCR1+ DC (migrated % of input, 6SEM) when tested with highly enriched (∼70%) splenic DC in vitro at 10, 100, and 1000 ng/ml (the indicated concentrations are calculated on the content of XCL1 in the reagents). (B) C57BL/6 mice were injected with OVA (2 mg), MARX10-OVA (0.16 mg), XCL1-OVA (1 mg; the indicated mass is calculated on the net content of OVA in the reagents), XCL1 (5 mg), PBS (negative control), or LPS (3 mg, positive control). Fourteen hours later, splenic DC (CD11c+ MHC II+ Lin2) were analyzed for expression levels of CD80, CD86, CD40, and PD-L2, shown as D mean ﬂuorescence intensity (MFI; difference between the MFI of the speciﬁc Abs and the MFI of the isotype controls). All experiments shown are representative of two independent experiments. 1074 SPECIFIC Ag TARGETING TO CROSS-PRESENTING DC

XCR1 is also processed in the MHC II presentation pathway, XCR1-deficient mice (B6.XCR1-lacZ+/+), only MARX10-OVA at C57BL/6 mice were adoptively transferred with naive OT-II 0.5 mg induced some CD4+ T cell proliferation (Fig. 4B). These CD4+ T cells and injected 1 d later with various doses of OVA, experiments demonstrated that Ags are specifically targeted to MARX10-OVA, or XCL1-OVA i.v., together with LPS. The fre- XCR1 and are then efficiently processed and presented in the quency of proliferating OT-II CD4+ T cells was determined 48 h context of MHC II. To determine whether the activated CD4+ later. T cell proliferation could already be observed with 0.005 mg T cells are also polarized in vivo, the transferred OT-II cells were MARX10-OVA (net content of OVA), and maximal proliferation analyzed for the production of various cytokines 72 h after transfer was reached with 0.05 mg (Fig. 4A); MARX10-OVA was thus by in vitro stimulation with PMA and ionomycin, followed by ∼1000-fold more efficient than nontargeted OVA in this assay intracellular staining. Targeting of Ag into XCR1+ DC in the (Fig. 4A). XCL1-OVA was ∼10-fold less efficient than MARX10- presence of adjuvant resulted in the secretion of IFN-g by a sub- OVA, but still highly effective as a targeting reagent. When these stantial proportion of Ag-specific CD4+ T cells (Fig. 4C), whereas reagents were tested for their targeting specificity by injection into IL-4 and IL-10 were not detected (our unpublished data). These Downloaded from http://www.jimmunol.org/ by guest on October 1, 2021

FIGURE 4. MARX10-OVA or XCL1-OVA triggers proliferation and Th1 polarization of CD4+ T cells in vivo and induces Ag-specific Ab. C57BL/6 mice were adoptively transferred with 1 3 106 naive, CFSE-labeled OT-II CD4+ T cells. One day later, recipients were injected i.v. with the indicated amounts of soluble OVA, MARX10-OVA, or XCL1-OVA (mass calculated on the net content of OVA) together with LPS (3 mg), and control animals received PBS only. (A) Proliferation of OT-II T cells was determined 48 h after injection of Ag by measuring the percentage of cells that have undergone at least one cell division. (B) Experiment was performed as in (A), but recipients were homozygous B6.XCR1-lacZ mice lacking XCR1. (C) Splenocytes were restimulated 72 h after immunization with PMA and ionomycin for 4.5 h in vitro in the presence of brefeldin A, and IFN-g was measured by intracellular flow cytometry; shown is the percentage of IFN g+ OT-II cells. (D) C57BL/6 mice were injected with OVA (2 mg), MARX-OVA (0.16 mg), or XCL1-OVA (1 mg, mass calculated on the net content of OVA) i.v., together with LPS (3 mg); control animals received LPS alone or PBS. Indicated levels of OVA- specific Ab isotypes were determined in serial dilutions of serum by ELISA on day 14 (4 mice per group; shown is the mean; maximal SEM was 0.08). All data shown are representative of at least two independent experiments. The Journal of Immunology 1075 results indicated that targeting of Ag via XCR1 in the presence of toxicity was measured 7 d after immunization using an in vivo an adjuvant induces Th1 polarization in Ag-specific CD4+ T cells. cytotoxicity assay. Injection of nontargeted OVA with LPS failed Finally, we tested whether the CD4+ T cells acquire helper to induce cytotoxicity over a large dose range of 30–1000 mg. functions and promote the generation of Ag-specific Abs. To this Only at the highest dose (3000 mg) did nontargeted OVA induce end, C57BL/6 mice were injected with OVA together with LPS relevant cytotoxicity (∼40% specific lysis; Fig. 6A). In stark i.v., either in a nontargeted form, or targeted as MARX10-OVA or contrast, already small amounts of MARX10-OVA (0.05 mg) in- XCL1-OVA. All three forms of Ag delivery induced substantial duced high killing activity (∼60%), and cytotoxicity levels of levels of IgM, IgG1, IgG2a, and IgG2b OVA-specific Abs, as ∼95–100% were reached by injection of 0.5 mg MARX10-OVA or determined in the serum 14 d later (Fig. 4D). more (Fig. 6A). Similarly, XCL1-OVA at a dose of 1 mg or more gave in vivo killing activity of up to 95–100% (Fig. 6A). When MARX10-OVA and XCL1-OVA efficiently induce proliferation any of the reagents were injected without LPS, no cytotoxicity was of CD8+ T cells in vivo observed (data not shown). To assess the targeting specificity also in Next, we determined the capacity of the two OVA-targeting sys- the cytotoxicity assay, analogous experiments were performed in + tems to induce CD8 T cell proliferation in vivo. C57BL/6 mice B6.XCR1-lacZ+/+ mice. In this study, injection of either MARX10- + were adoptively transferred with resting OT-I CD8 T cells and OVA or XCL1-OVA at the higher dose range did not give any sig- 1 d later injected i.v. with various doses of OVA, MARX10-OVA, nificant cytotoxicity (Fig. 6B). These experiments determined that + or XCL1-OVA, together with LPS. The proportion of OT-I CD8 targeting of Ag into DC via XCR1 is highly effective in inducing T cells that have undergone at least one cell division was deter- CD8+ T cell cytotoxicity, whereas injection of nontargeted OVA is + mined 48 h later. Minimal CD8 T cell proliferation could already essentially ineffective, even at very high amounts. be observed after application of 0.0016 mg MARX10-OVA, and Downloaded from the maximal signal was reached with ∼0.016–0.05 mg (Fig. 5A). Vaccination with MARX10-OVA or XCL1-OVA prevents tumor MARX10-OVA was thus at least 200-fold more effective than outgrowth nontargeted OVA (Fig. 5A). XCL1-OVA was ∼20-fold less effi- To determine whether the cytotoxicity induced through targeting of cient than MARX10-OVA, but still ∼10-fold more potent than OVA into XCR1+ DC translates into tumoricidal activity in vivo, nontargeted OVA in this assay. The specificity of Ag targeting was naive mice were immunized with MARX10-OVA, XCL1-OVA, or assessed by injection of the reagents into homozygous B6.XCR1- nontargeted OVA, together with LPS. Seven days later, all mice http://www.jimmunol.org/ lacZ mice. In this study, only higher amounts of MARX10-OVA were injected s.c. with a high number (6 3 105) of E.G7 cells, an (0.16 mg) induced proliferation of a proportion of CD8+ T cells, aggressively growing syngeneic EL4 thymoma transfected with and thus had a similar effect as nontargeted OVA (Fig. 5B). OVA. In control animals and mice injected with nontargeted OVA, Together, these experiments once more demonstrated that both the tumor grew quickly to finally reach a size of ∼1 cm in di- MARX10-OVA and XCL1-OVA specifically target Ag into ameter 14 d after inoculation. In clear contrast, all animals im- XCR1+ DC. At the same time, the results indicated that Ags tar- munized with the targeted forms of OVA effectively controlled geted to XCR1 are efficiently shunted into the MHC I cross- tumor outgrowth (Fig. 6C). This experiment determined that the presentation pathway. induced cytotoxicity is of biological relevance. by guest on October 1, 2021 MARX10-OVA and XCL1-OVA are highly effective in inducing Targeting of Ag via human XCR1 induces potent cytotoxicity CD8+ T cell cytotoxicity in vivo in vivo Induction of CD8+ T cell cytotoxicity was tested in naive mice, Having demonstrated the efficiency of Ag targeting via XCR1 in without an adoptive transfer of T cells. C57BL/6 mice were the mouse, we sought to determine whether this approach also injected once with the indicated amounts of OVA, MARX10-OVA, functions using human XCL1 and its variant XCL2 (differing only or XCL1-OVA, together with LPS. The degree of induced cyto- at positions 28 and 29) to target human XCR1 under physiological

FIGURE 5. MARX10-OVA or XCL1-OVA induces proliferation of CD8+ T cells in vivo. (A) C57BL/6 mice were adoptively transferred with 2 3 106 resting, CFSE-labeled OT-I CD8+ T cells. One day after transfer, recipient animals were injected i.v. with the indicated amounts of soluble OVA, MARX10- OVA, or XCL1-OVA (mass calculated on the net content of OVA) in the presence of LPS (3 mg); control animals received PBS. Proliferation of OT-I T cells was determined 48 h after Ag injection by measuring the percentage of cells that have undergone at least one cell division. (B) Experiment was performed as in (A), but recipients were homozygous B6.XCR1-lacZ animals lacking XCR1. The data shown are representative of two independent experiments. 1076 SPECIFIC Ag TARGETING TO CROSS-PRESENTING DC

FIGURE 6. MARX10-OVA and XCL1-OVA are highly effective in inducing CD8+ T cell cytotoxicity in vivo. (A)NaiveC57BL/6mice were immunized i.v. once with the indicated amounts of OVA, MARX10- OVA, or XCL1-OVA (mass calculated on the net content of OVA), together with LPS (3 mg). On day 6, the animals were injected with CFSE-labeled target cells to quantitate the induced cytotoxicity in vivo (for details, see Materials and Methods). (B) The experiment was performed as in (A), but treated animals were homozygous B6.XCR1-lacZ mice lacking XCR1. (C) Naive C57BL/6 mice were im-

munized once with MARX10-OVA Downloaded from (0.16 mg), XCL1-OVA (1 mg), or OVA (1 mg), together with LPS (3 mg) i.v., and control mice were only injected with LPS. Seven days later, mice were inoculated s.c. with 6 3 105 E.G7 cells. On day 14, mice were sacriﬁced and the weight of the tumor http://www.jimmunol.org/ tissue was determined. The data shown are representative of three (A and B) and two (C) independent experiments.

conditions. To this end, we constructed and expressed huXCL1- CD8+ T cells allowed to define ideal criteria for the design and by guest on October 1, 2021 StrepTag, huXCL2-StrepTag, huXCL1-OVA, and huXCL2-OVA evaluation of novel vaccines inducing a cytotoxic response. Such (Fig. 7A). At the same time, we generated mice selectively express- a vaccine should do the following: 1) deliver Ags into cross- ing human XCR1 on cross-presenting DC on a background deficient presenting DC in a highly specific fashion; 2) have no adverse for murine XCR1 (huXCR1tg). In a first step, binding of the various effects on the targeted DC or other cells; 3) allow processing of reagents to human and murine XCR1 was determined. huXCL1 the delivered Ags in the MHC I pathway; 4) be devoid of any stained 55–80% of CD141+ DC obtained from PBMC or human inherently immunogenic components; 5) have a high tissue pen- spleens, but not other DC (Fig. 7B), or other cells (data not shown). etration; 6) provide signals for the differentiation of CD8+ T cells Important for subsequent functional experiments, huXCL1 bound to to cytotoxic effectors; and 7) be applicable in humans. We would human XCR1 in huXCR1tg mice, but did not react with murine XCR1 like to discuss the data presented in this work in view of these in C57BL/6 controls (Fig. 7B). The binding pattern of huXCL2 was criteria. indistinguishable from the pattern observed with huXCL1 (Fig. 7B). muXCL1 cross-reacted with human XCR1 in PBMC, spleens, and the Targeting specificity huXCR1tg mouse, in which it also detected a XCR1lowMHChigh DC In the mouse, the chemokine receptor XCR1 is exclusively ex- population (Fig. 7B). The targeting constructs containing full-length pressed on cross-presenting DC throughout the body. XCR1 is, in OVA (huXCL1-OVA, huXCL2-OVA) exhibited a staining pattern fact, the lineage marker for cross-presenting DC in the mouse (19, comparable to huXCL1 and huXCL2 (our unpublished data). 28). This restricted expression is in the human mirrored by the To assess the capacity of human XCL1 or XCL2 to target Ag exclusive presence of XCR1 on the surface of CD141+ DC, the in vivo, huXCR1tg mice were immunized once with varying counterpart of murine cross-presenting DC (18, 20, 21, 29). Al- amounts of XCL1-OVA or XCL2-OVA i.v., together with LPS. The though in vivo tests of Ag cross-presentation are not possible in in vivo cytotoxicity assay on day 6 revealed that 1 mg huXCL1- humans, the highly conserved expression and function of XCR1 OVA or 2 mg huXCL2-OVA induced maximal killing activity of and its chemokine ligand XCL1/2 strongly indicate that targeting 95–100% (Fig. 7C), whereas the same amounts were largely in- of Ag via XCR1 should lead to efficient cross-presentation also in effective in wild-type C57BL/6 mice (Fig. 7D). This experiment humans. demonstrated that human XCL1 or XCL2 is capable of efficiently For the XCR1-specific mAb MARX10, the targeting specificity targeting an Ag into cross-presenting DC expressing human XCR1 could directly be visualized in vivo. As long as injected i.v. at lower, under physiological conditions. but saturating amounts (1–2 mg), the mAb selectively bound to XCR1+ DC. Fusing the model Ag OVA to either mAb MARX10 Discussion or XCL1 did not change their specificity for XCR1. The speci- In recent years, the growing understanding of the role of DC subsets ficity of Ag delivery could also be documented for both vector in the uptake, processing, and presentation of Ag to CD4+ and systems at a functional level with mice deficient for XCR1. The Journal of Immunology 1077 Downloaded from http://www.jimmunol.org/ by guest on October 1, 2021

FIGURE 7. Human XCR1 can be targeted for induction of potent cytotoxic immunity in vivo. (A) Recombinant fusion proteins huXCL1, huXCL2, huXCL1-OVA, and huXCL2-OVA (drawn to scale). (B) Binding of Strep-tagged huXCL1, huXCL2, and muXCL1 to DC (which were deﬁned both in the human and mouse systems as CD11c+ MHC II+ Lin2) in 1) human PBMC, 2) human spleen, 3) spleens of transgenic mice expressing only human XCR1 on cross-presenting DC (huXCR1tg), and 4) spleens of wild-type C57BL/6 mice; background control (gray histograms) was with Strep-tactin only. (C) Naive huXCR1tg or (D) wild-type C57BL/6 were immunized i.v. with the indicated amounts of huXCL1-OVA or XCL2-OVA (mass based on net content of OVA) together with 3 mg LPS. On day 6, the animals were injected with CFSE-labeled target cells to quantitate the induced cytotoxicity in vivo. The data shown are representative of two independent experiments.

Currently, only one other molecule is known to be similarly phagocytosed necrotic cells and thus to be an integral part of the restricted to cross-presenting DC. In humans, Clec9A/DNGR-1 is cross-presentation machinery (33–36), as is XCR1 (17). selectively expressed on CD141+ DC (30, 31); the only report When comparing mAb MARX10 with XCL1 for targeting of Ag, indicating some presence of Clec9A/DNGR-1 on other cell types it is apparent that MARX10 binds to XCR1 with higher avidity. (32) has to be independently veriﬁed yet. Interestingly, the lectin Moreover, MARX10 can be detected on the surface of cross- Clec9A/DNGR-1 has been shown to optimize processing of presenting DC still 1 wk after injection, suggesting a relatively 1078 SPECIFIC Ag TARGETING TO CROSS-PRESENTING DC

long t1/2 in the circulation. Given the higher avidity and the long in LN after systemic injection points toward this direction; an engi- t1/2, one would expect that the mAb clearly outperforms the neered smaller Ab format could overcome these potential limitations. chemokine ligand in the delivery of Ag to XCR1. However, when XCL1/2, in contrast, is a small protein of ∼10 kDa. As such, this tested for induction of CD8+ T cell cytotoxicity, the difference in molecule can be expected to have a high tissue penetration, which efficiency between those two vector systems was surprisingly may only be limited by the size of the fused protein; in the case of small in a large series of experiments. Why XCL1 is so efficient in (long) peptides, the tissue penetration should remain high. delivering Ag into XCR1+ DC remains unclear at present. Induction of CD8+ T cell cytotoxicity No adverse effects on targeted DC or other cells Upon targeting of Ag, we could induce potent CD8+ cytotoxicity The use of the nonblocking, nonagonistic mAb MARX10 dem- when applied together with the adjuvant LPS; in the absence of an onstrates that delivery of Ag via XCR1 using a mAb can be done adjuvant, only proliferation resulted. This finding is in accordance without activation or depletion of the targeted DC. The chemokine with other targeting systems (5–8) and indicates that inflammatory ligand XCL1 does elicit a short Ca2+ burst upon binding to XCR1 signals essential for induction of cytotoxicity have to be provided. and is an effective chemoattractant (17). Nonetheless, application In the present work, we used LPS throughout the study, but ad- of high amounts of XCL1 does not induce a sustained activation of ditional experiments indicate that also other TLR agonists such as DC in vivo, as determined by measuring activation surface mol- CpG or poly(I:C) are fully effective (our unpublished data). Be- ecules. Similarly, no adverse effects of the vector systems are to be cause TLR4, the receptor for LPS, is not present in human cross- expected on nontarget cells, because Abs (as a protein class) and presenting DC (21), other adjuvants such as poly(I:C) will have to the chemokine XCL1/2 are physiological proteins present in the be applied in the human. Downloaded from circulation. Applicability in humans Processing and presentation of delivered Ag in the MHC I We could demonstrate that human XCL1/2-OVA induces potent pathway CD8+ cytotoxicity when injected into mice expressing only human Upon binding of their ligands, chemokine receptors are classically XCR1 on their cross-presenting DC. This finding demonstrates internalized. Therefore, it is not surprising that Ag bound to XCL1, that huXCL1/2 in vivo binds to human XCR1 with sufficient when used as a vector, is efficiently taken up by DC. Interestingly, avidity to be used as a targeting system. Furthermore, these results http://www.jimmunol.org/ this is apparently also the case with mAb MARX10, which has no indicate a rapid internalization of human XCR1, followed by an chemokine agonist activity. Because the biochemical pathways efficient Ag processing and presentation in the MHC I pathway. involved in cross-presentation of Ag are not well defined, pre- Because one can assume that cross-presenting human DC process sentation in the MHC I pathway has to be determined empirically. and present Ag with the same efficiency as their mouse counter- In this respect, XCR1 as a target structure appears to be highly parts, these results strongly indicate that targeting of Ag to XCR1 effective, because very low amounts of MARX10-OVA or XCL1- is a viable approach for the induction of cytotoxicity in humans. In OVA are capable of inducing potent CD8+ cytotoxicity. the next step, this has to be demonstrated in nonhuman primates. In addition to their capacity for Ag cross-presentation, Batf3- by guest on October 1, 2021 dependent DC can also classically process and present Ag in the Acknowledgments MHC II pathway. In our experiments, we have observed that We thank Michel C. Nussenzweig for the generous gift of the cDNA of mAb targeting of OVAvia XCR1 quite efficiently induces activation (our DEC-205. unpublished data), proliferation, and Th1 differentiation of CD4+ T cells, and leads to the generation of Ag-specific Abs. This Disclosures indicates that Ag internalized by XCR1 is shunted both to the R.A.K. is inventor of a patent application by the Robert Koch-Institute on MHC I and MHC II pathways, similar to Ag internalized by targeting XCR1 for vaccination. Clec9A/DNGR-1 (30, 31). Absence of inherently immunogenic components References 1. Rappuoli, R., and A. Aderem. 2011. A 2020 vision for vaccines against HIV, Vaccines usually have to be applied multiple times. It is, therefore, tuberculosis and malaria. Nature 473: 463–469. of major importance whether the targeting vector (without attached 2. Jenkins, M. K., H. H. Chu, J. B. McLachlan, and J. J. Moon. 2010. On the composition of the preimmune repertoire of T cells specific for peptide-major Ag) is immunogenic. For example, viral vectors typically induce histocompatibility complex ligands. Annu. Rev. Immunol. 28: 275–294. strong immune reactions and memory, and therefore cannot be 3. Hawiger, D., K. Inaba, Y. Dorsett, M. Guo, K. Mahnke, M. Rivera, J. V. Ravetch, applied repeatedly. Ample clinical experience indicates that tar- R. M. Steinman, and M. C. Nussenzweig. 2001. Dendritic cells induce peripheral T cell unresponsiveness under steady state conditions in vivo. J. Exp. Med. 194: geting of Ag by humanized or human mAb should not pose such 769–779. problems. Human XCL1/2 is a physiological protein abundantly se- 4. Bonifaz, L., D. Bonnyay, K. Mahnke, M. Rivera, M. C. Nussenzweig, and R. M. Steinman. 2002. Efficient targeting of protein antigen to the dendritic cell creted into the circulation upon activation of the innate immune system receptor DEC-205 in the steady state leads to antigen presentation on major + andbyactivatedCD8 T cells, and therefore tolerance to this protein histocompatibility complex class I products and peripheral CD8+ T cell toler- is fully established. Thus, repeated application of either vector system ance. J. Exp. Med. 196: 1627–1638. + 5. Kreutz, M., P. J. Tacken, and C. G. Figdor. 2013. Targeting dendritic cells: why fortargetingofAgintoXCR1 DC should be effective. This may be bother? Blood 121: 2836–2844. quite important in the successive delivery of different Ags using 6. Palucka, K., and J. Banchereau. 2013. Dendritic-cell-based therapeutic cancer therapeutic vaccination in neoplastic diseases. vaccines. Immunity 39: 38–48. 7. Radford, K. J., and I. Caminschi. 2013. New generation of dendritic cell vaccines. Hum. Vaccin. Immunother. 9: 259–264. Tissue penetration 8. Kastenmuller,€ W., K. Kastenmuller,€ C. Kurts, and R. A. Seder. 2014. Dendritic cell targeted vaccines - hope or hype? Nat. Rev. Immunol. 14: 705–711. The relatively high m.w. of Abs prevents their rapid clearance from 9. den Haan, J. M., S. M. Lehar, and M. J. Bevan. 2000. CD8+ but not CD82 the circulation, when systemically applied. However, this high m.w. dendritic cells cross-prime cytotoxic T cells in vivo. J. Exp. Med. 192: 1685– may compromise the tissue penetration of Ab targeting vectors in 1696. 10. Pooley, J. L., W. R. Heath, and K. Shortman. 2001. Cutting edge: intravenous peripheral lymphoid organs. Our finding of a rather selective de- soluble antigen is presented to CD4 T cells by CD82 dendritic cells, but cross- tection of MARX10 on resident DC (as opposed to migratory DC) presented to CD8 T cells by CD8+ dendritic cells. J. Immunol. 166: 5327–5330. The Journal of Immunology 1079

11. Hildner, K., B. T. Edelson, W. E. Purtha, M. Diamond, H. Matsushita, 24. Schneider, I. 1972. Cell lines derived from late embryonic stages of Drosophila M. Kohyama, B. Calderon, B. U. Schraml, E. R. Unanue, M. S. Diamond, et al. melanogaster. J. Embryol. Exp. Morphol. 27: 353–365. 2008. Batf3 deficiency reveals a critical role for CD8a+ dendritic cells in cy- 25. Lo¨hning, M., A. Hutloff, T. Kallinich, H. W. Mages, K. Bonhagen, A. Radbruch, totoxic T cell immunity. Science 322: 1097–1100. E. Hamelmann, and R. A. Kroczek. 2003. Expression of ICOS in vivo defines 12. Hashimoto, D., J. Miller, and M. Merad. 2011. Dendritic cell and macrophage CD4+ effector T cells with high inflammatory potential and a strong bias for heterogeneity in vivo. Immunity 35: 323–335. secretion of interleukin 10. J. Exp. Med. 197: 181–193. 13. Dudziak, D., A. O. Kamphorst, G. F. Heidkamp, V. R. Buchholz, 26. Flavell, D. J., S. U. Flavell, D. A. Boehm, L. Emery, A. Noss, N. R. Ling, C. Trumpfheller, S. Yamazaki, C. Cheong, K. Liu, H. W. Lee, C. G. Park, et al. P. R. Richardson, D. Hardie, and D. H. Wright. 1995. Preclinical studies with the 2007. Differential antigen processing by dendritic cell subsets in vivo. Science anti-CD19-saporin immunotoxin BU12-SAPORIN for the treatment of human- 315: 107–111. B-cell tumours. Br. J. Cancer 72: 1373–1379. 14. Inaba, K., W. J. Swiggard, M. Inaba, J. Meltzer, A. Mirza, T. Sasagawa, 27. Moore, M. W., F. R. Carbone, and M. J. Bevan. 1988. Introduction of soluble M. C. Nussenzweig, and R. M. Steinman. 1995. Tissue distribution of the DEC- protein into the class I pathway of antigen processing and presentation. Cell 54: 205 protein that is detected by the monoclonal antibody NLDC-145. I. Ex- 777–785. € pression on dendritic cells and other subsets of mouse leukocytes. Cell. Immunol. 28. Becker, M., S. Guttler, A. Bachem, E. Hartung, A. Mora, A. Ja¨kel, A. Hutloff, V. Henn, H. W. Mages, S. Gurka, and R. A. Kroczek. 2014. Ontogenic, phe- 163: 148–156. + 15. Witmer-Pack, M. D., W. J. Swiggard, A. Mirza, K. Inaba, and R. M. Steinman. notypic, and functional characterization of XCR1 dendritic cells leads to 1995. Tissue distribution of the DEC-205 protein that is detected by the a consistent classification of intestinal dendritic cells based on the expression of monoclonal antibody NLDC-145. II. Expression in situ in lymphoid and non- XCR1 and SIRPa. Front. Immunol. 5: 326. lymphoid tissues. Cell. Immunol. 163: 157–162. 29. Poulin, L. F., M. Salio, E. Griessinger, F. Anjos-Afonso, L. Craciun, J.-L. Chen, A. M. Keller, O. Joffre, S. Zelenay, and E. Nye, et al. 2010. Characterization of 16. Kato, M., K. J. McDonald, S. Khan, I. L. Ross, S. Vuckovic, K. Chen, + + + D. Munster, K. P. MacDonald, and D. N. Hart. 2006. Expression of human DEC- human DNGR-1 BDCA3 leukocytes as putative equivalents of mouse CD8a dendritic cells. J. Exp. Med. 207: 1261–1271. 205 (CD205) multilectin receptor on leukocytes. Int. Immunol. 18: 857–869. 30. Caminschi, I., A. I. Proietto, F. Ahmet, S. Kitsoulis, J. Shin Teh, J. C. Lo, 17. Dorner, B. G., M. B. Dorner, X. Zhou, C. Opitz, A. Mora, S. Guttler,€ A. Hutloff, A. Rizzitelli, L. Wu, D. Vremec, S. L. van Dommelen, et al. 2008. The dendritic H. W. Mages, K. Ranke, M. Schaefer, et al. 2009. Selective expression of the cell subtype-restricted C-type lectin Clec9A is a target for vaccine enhancement. chemokine receptor XCR1 on cross-presenting dendritic cells determines co- + Blood 112: 3264–3273. operation with CD8 T cells. Immunity 31: 823–833. Downloaded from 31. Sancho, D., D. Mourao-Sa´, O. P. Joffre, O. Schulz, N. C. Rogers, 18. Crozat, K., R. Guiton, V. Contreras, V. Feuillet, C.-A. Dutertre, E. Ventre, D. J. Pennington, J. R. Carlyle, and C. Reis e Sousa. 2008. Tumor therapy in T.-P. V. Manh, T. Baranek, A. K. Storset, J. Marvel, et al. 2010. The XC che- mice via antigen targeting to a novel, DC-restricted C-type lectin. J. Clin. Invest. mokine receptor 1 is a conserved selective marker of mammalian cells homol- 118: 2098–2110. ogous to mouse CD8a+ dendritic cells. J. Exp. Med. 207: 1283–1292. € 32. Huysamen, C., J. A. Willment, K. M. Dennehy, and G. D. Brown. 2008. 19. Bachem, A., E. Hartung, S. Guttler, A. Mora, X. Zhou, A. Hegemann, CLEC9A is a novel activation C-type lectin-like receptor expressed on BDCA3+ M. Plantinga, E. Mazzini, P. Stoitzner, S. Gurka, et al. 2012. Expression of dendritic cells and a subset of monocytes. J. Biol. Chem. 283: 16693–16701. XCR1 characterizes the Batf3-dependent lineage of dendritic cells capable of 33. Ahrens, S., S. Zelenay, D. Sancho, P. Hanc, S. Kjaer, C. Feest, G. Fletcher, antigen cross-presentation. Front. Immunol. 3: 214. C. Durkin, A. Postigo, M. Skehel, et al. 2012. F-actin is an evolutionarily http://www.jimmunol.org/ € 20. Bachem, A., S. Guttler, E. Hartung, F. Ebstein, M. Schaefer, A. Tannert, conserved damage-associated molecular pattern recognized by DNGR-1, a re- A. Salama, K. Movassaghi, C. Opitz, H. W. Mages, et al. 2010. Superior antigen ceptor for dead cells. Immunity 36: 635–645. + + cross-presentation and XCR1 expression define human CD11c CD141 cells as 34. Iborra, S., H. M. Izquierdo, M. Martinez-Lopez, N. Blanco-Menendez, C. Reis e + homologues of mouse CD8 dendritic cells. J. Exp. Med. 207: 1273–1281. Sousa, and D. Sancho. 2012. The DC receptor DNGR-1 mediates cross-priming 21. Jongbloed, S. L., A. J. Kassianos, K. J. McDonald, G. J. Clark, X. Ju, C. E. Angel, of CTLs during vaccinia virus infection in mice. J. Clin. Invest. 122: 1628–1643. + C.-J. J. Chen, P. R. Dunbar, R. B. Wadley, V. Jeet, et al. 2010. Human CD141 35. Zelenay, S., A. M. Keller, P. G. Whitney, B. U. Schraml, S. Deddouche, + (BDCA-3) dendritic cells (DCs) represent a unique myeloid DC subset that N. C. Rogers, O. Schulz, D. Sancho, and C. Reis e Sousa. 2012. The dendritic cross-presents necrotic cell antigens. J. Exp. Med. 207: 1247–1260. cell receptor DNGR-1 controls endocytic handling of necrotic cell antigens to 22. Rostovskaya, M., R. Naumann, J. Fu, M. Obst, D. Mueller, A. F. Stewart, and favor cross-priming of CTLs in virus-infected mice. J. Clin. Invest. 122: 1615– K. Anastassiadis. 2013. Transposon mediated BAC transgenesis via pronuclear 1627. injection of mouse zygotes. Genesis 51: 135–141. 36. Zhang, J. G., P. E. Czabotar, A. N. Policheni, I. Caminschi, S. S. Wan,

23. Wallny, H. J. 1997. Production of soluble MHC class II molecules in Drosophila S. Kitsoulis, K. M. Tullett, A. Y. Robin, R. Brammananth, M. F. van Delft, et al. by guest on October 1, 2021 melanogaster Schneider cells. In Immunological Methods Manual, Vol. 1. 2012. The dendritic cell receptor Clec9A binds damaged cells via exposed actin I. Lefkovits, ed., Academic Press, San Diego, CA, p. 51–59. ﬁlaments. Immunity 36: 646–657.