Tumors Anti-DR5 Antibody Therapy in Established Contribute to The

CD11c+ Dendritic Cells and B Cells Contribute to the Tumoricidal Activity of Anti-DR5 Antibody Therapy in Established Tumors This information is current as of September 28, 2021. Nicole M. Haynes, Edwin D. Hawkins, Ming Li, Nicole M. McLaughlin, Günter J. Hämmerling, Reto Schwendener, Astar Winoto, Allen Wensky, Hideo Yagita, Kazuyoshi Takeda, Michael H. Kershaw, Phillip K. Darcy and Mark J. Smyth Downloaded from J Immunol published online 26 May 2010 http://www.jimmunol.org/content/early/2010/05/26/jimmun ol.0903624 http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication by guest on September 28, 2021

*average

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

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Published May 26, 2010, doi:10.4049/jimmunol.0903624 The Journal of Immunology

CD11c+ Dendritic Cells and B Cells Contribute to the Tumoricidal Activity of Anti-DR5 Antibody Therapy in Established Tumors

Nicole M. Haynes,* Edwin D. Hawkins,* Ming Li,* Nicole M. McLaughlin,* Gu¨nter J. Ha¨mmerling,† Reto Schwendener,‡ Astar Winoto,x Allen Wensky,x Hideo Yagita,{ Kazuyoshi Takeda,{ Michael H. Kershaw,*,‖ Phillip K. Darcy,*,‖ and Mark J. Smyth*,‖

The selective targeting of the tumor-associated death-inducing receptors DR4 and DR5 with agonistic mAbs has demonstrated preclinical and clinical antitumor activity. However, the cellular and molecular mechanisms contributing to this efficacy remain Downloaded from poorly understood. In this study, using the first described C57BL/6 (B6) TRAIL-sensitive experimental tumor models, we have characterized the innate and adaptive immune components involved in the primary rejection phase of an anti-mouse DR5 (mDR5) mAb, MD5-1 in established MC38 colon adenocarcinomas. FcR mediated cross-linking of MD5-1 significantly inhibited the growth of MC38 colon adenocarcinomas through the induction of TRAIL-R–dependent tumor cell apoptosis. The loss of host DR5, TRAIL, perforin, FasL, or TNF did not compromise anti-DR5 therapy in vivo. By contrast, anti-DR5 therapy was completely abrogated in mice deficient of B cells or CD11c+ dendritic cells (DCs), providing the first direct evidence that these cells http://www.jimmunol.org/ play a critical role. Importantly, the requirement for an intact B cell compartment for optimal anti-DR5 antitumor efficacy was also observed in established AT-3 mammary tumors. Interestingly, MD5-1–mediated apoptosis as measured by early TUNEL activity was completely lost in B cell-deficient mMT mice, but intact in mice deficient in CD11c+ DCs. Overall, these data show that Ab-mediated targeting of DR5 triggers tumor cell apoptosis in established tumors in a B cell-dependent manner and that CD11c+ DCs make a critical downstream contribution to anti-DR5 antitumor activity. The Journal of Immunology, 2010, 185: 000–000.

he therapeutic targeting of tumors or components of the processes, such as proliferation, differentiation, and/or survival. immune system with molecule-specific mAbs is now con- Through Fc-mediated Ab-dependent cellular cytotoxicity (ADCC), T sidered a viable treatment option for cancer patients. Tras- complement-dependent cytotoxicity, opsonisation, and/or phagocy- by guest on September 28, 2021 tuzumab (Herceptin), which binds the breast cancer-associated tosis mechanisms, mAbs may also trigger tumor cell death resulting growth factor receptor ErbB2, and rituximab (Rituxan) that targets in the liberation of tumor-associated Ags in a manner that can pro- B cells for elimination by binding the B cell-associated marker mote tumor immunity (3–5). These cell death-inducing and im- CD20, are two such prototypic Abs that are currently in clinical mune-stimulating properties of mAbs have made them valuable use. Indeed, these Abs have been used successfully in combination therapeutic tools in cancer treatment and with current advances in with chemotherapy for the treatment of patients with breast cancer the fields of immunology and oncology, novel molecular targets are and non-Hodgkin’s lymphoma, respectively (1, 2). Tumor- being elucidated that will significantly broaden the therapeutic ap- targeting mAbs exert their antitumor effects by modulating basic plication of mAb-based immunotherapies. Tumor-targeted mAbs that have generated a considerable amount of interest in oncology are those that bind the receptors for a death- *Cancer Immunology Program, Trescowthick Laboratories, Peter MacCallum Cancer Centre, East Melbourne; ‖Department of Pathology, University of Melbourne, Parkville, inducing ligand, TRAIL (6, 7). Unlike other death-inducing ligands, Victoria, Australia; †Division of Molecular Immunology, German Cancer Research TRAIL preferentially kills malignant cells via activating the death Center (Deutsches Krebsforschungszentrum), Heidelberg, Germany; ‡Institute of Mo- x receptors DR4 and DR5, while sparing most normal cells. How- lecular Cancer Research, University of Zurich, Zurich, Switzerland; Cancer Research Laboratory, Department of Molecular and Cell Biology, University of California, Ber- ever, although many tumors express TRAIL death receptors, they keley, Berkeley, CA 94720; and {Department of Immunology, Juntendo University also typically express TRAIL decoy receptors that have the po- School of Medicine, Tokyo, Japan tential to limit the therapeutic efficacy of recombinant TRAIL (8). Received for publication November 9, 2009. Accepted for publication April 20, 2010. MAbs that specifically bind the death receptors of TRAIL bypass This work was supported by a National Health and Medical Research Council of this problem, and offer the advantage of having a longer half-life Australia program grant, a Cancer Council of Victoria project grant, and a Cure Cancer Australia grant. M.J.S. was supported by a National Health and Medical in vivo (9), higher affinity for the target receptor, and the potential Research Council of Australia fellowship. N.M.H was supported in part by a National to promote the recruitment and activation of immune effector cells Health and Medical Research Council of Australia C.J. Martin fellowship. (10). Furthermore, selective targeting of the TRAIL receptors sig- Address correspondence and reprint requests to Dr. Nicole Haynes or Prof. Mark nals cell death in the absence of p53 function; the loss of which Smyth, Cancer Immunology Program, Peter MacCallum Cancer Centre, Locked Bag 1, A’Beckett Street, Melbourne, Victoria 8006, Australia. E-mail addresses: has been shown to disable chemotherapeutic drug responses (11). [email protected] or [email protected] With encouraging preclinical data and validation that the TRAIL/ Abbreviations used in this paper: ADCC, Ab-dependent cellular cytotoxicity; cIg, TRAIL-receptor pathway is important in tumor immunosurveil- control Ig; DC, dendritic cell; DT, diphtheria toxin; MHC I, MHC class I; mTRAIL, lance (12–14), the therapeutic efficacy and toxicity of agonistic mouse TRAIL; pfp, perforin; WT, wild-type. mAbs to human DR4 and DR5 are now being evaluated in early Copyright Ó 2010 by The American Association of Immunologists, Inc. 0022-1767/10/$16.00 phase clinical trials, alone and in combination with first-line

www.jimmunol.org/cgi/doi/10.4049/jimmunol.0903624 2 TUMORICIDAL MECHANISMS OF ANTI-DR5 mAb IN ESTABLISHED CANCER chemotherapies and irradiation (15). Results from these trials have supplemented with 10% heat-inactivated FCS (Moregate Biotech, Bulimba, indicated that these anti-TRAIL receptor mAbs can be adminis- Queensland, Australia), 100 mg/ml penicillin/streptomycin (Sigma-Aldrich, tered safely and are capable of promoting disease stabilization Castle Hill, New South Wales, Australia), and 2 mM L-glutamine (JRH Biosciences, Brooklyn, Victoria, Australia). The BALB/c-derived 4T1.2 and/or partial responses without apparent toxicity (15). Currently, mammary carcinoma has been previously described (18) and was cultured however, there is scant clinical data available on the impact of in RPMI 1640 media containing the aforementioned supplements. The these therapeutic mAbs on immune parameters. MC38-c-FLIP line was generated using retroviral gene-transfer technology Because the preclinical development of anti-DR4 and -DR5 ag- as previously described (18). Briefly, a standard calcium phosphate trans- fection method was used to introduce the murine stem cell virus-IRES-GFP onistic mAbs for cancer therapy has largely involved examining the containing the FLIPLong encoding sequence (kindly provided by Dr. Thomas tumoricidal activity of humanized reagents in immune compro- Sayers, National Cancer Institute, Frederick, MD) and the amphotrophic mised xenogeneic mouse models of human cancer, the immunolog- helper plasmid into the 293T packaging cell line. Viral supernatant collected ical consequences of anti-DR5 therapy on the host and tumor from these 293T cell cultures, were supplemented with 4 mg/ml polybrene microenvironment have remained poorly defined. Previously we (Millipore, Billerica, MA) and used to transduce the MC38 tumor cell line. Seventy-two hours after transduction, GFP-positive tumor cells were selected demonstrated an anti-mDR5 mAb (MD5-1) could exhibit potent for by flow cytometry-mediated cell sorting (Vantage-DIVA; BD Bioscien- antitumor effects against TRAIL-sensitive mouse tumors in synge- ces, San Jose, CA). neic BALB/c mice through FcgR-dependent induction of tumor cell apoptosis (10). When administered at the time of 4T1.2 mam- DR5 and MHC class I expression mary tumor cell implantation, anti-DR5 (MD5.1) was found to Cell surface DR5 and MHC class I (MHC I) levels were assessed by flow promote the recruitment of dendritic cells (DCs) and stimulate cytometry. Staining of cells involved treating 106 cells with biotinylated adaptive tumor immunity that could be raised by secondary anti-mouse DR5 (MD5-1; eBioscience, San Diego, CA), anti-mouse Downloaded from b b rechallenge (10). In this study, we have developed C57BL/6 H-2K (AF6-88.5; BD Pharmingen, San Jose, CA) or anti-mouse I-A TRAIL-sensitive tumor models with the aim of being able to (KH74; BD Pharmingen) in the presence of normal mouse serum (Jackson Immunoresearch Laboratory, West Grove, PA). Cells were subsequently use gene-targeted mice to better evaluate the innate and adaptive treated with allophycocyanin-conjugated streptavidin (eBioscience). Dur- immune cells contributing to the tumoricidal activity of the ing this staining procedure, tumor cells were incubated on ice and washed MD5-1 mAb in more clinically relevant established tumors. and resuspended in FACS buffer (2% FCS in PBS with 0.1% sodium azide and 1 mM EDTA) prior to being examined on a LSRII analyzer (BD C57BL/6 gene-targeted or immune cell-depleted mice were used http://www.jimmunol.org/ Biosciences). Cell viability was assessed through the addition of 100 to examine the antitumor activity of MD5-1 against the TRAIL- ng/ml DAPI (49,6-diamido-2-phenylindole; Invitrogen, Eugene, OR). sensitive mouse MC38 colon adenocarcinoma and AT-3 mammary carcinoma. In these tumor models, an intact B cell compartment Abs and reagents was critical for the therapeutic activity of MD5-1 against established tumors. B cells were confirmed to trigger tumor cell apo- Agonistic mAbs to mouse DR5 (MD5-1), neutralizing/depleting Abs to asialoGM1 (NK cell depletion; Wako Pure Chemicals, Osaka, Japan), ptosis by FcR-mediated cross-linking of the MD5-1 mAb in vitro mouse CD11b (5C6), mouse CD8a (53.6.7), and control Ig (cIg) hamster and in vivo B cells were critical for directly triggering MD5-1– (UC8-1B9) or rat (MAC4) Abs were prepared as described (10). For tumor mediated tumor cell apoptosis. By contrast, although DC subsets therapy, 100 mg MD5-1 or UC8-1B9 cIg were administered into the peri- could also trigger tumor cell apoptosis in vitro, CD11c+ DCs were toneal cavity (i.p.) of mice once every 4 d for a total of three treatments. In by guest on September 28, 2021 experiments using mMT and RAG-12/2 mice, 50 mg MD5-1 or UC8-1B9 not necessary for MD5-1–mediated apoptosis in vivo, but were cIg was used to limit the onset of toxic side effects in the immune com- important downstream for the antitumor activity of MD5-1. These promised mice. Neutralizing/depleting Abs to CD11b (300 mg), NK cells data are intriguing and establish an important platform on which (asialoGM1; 50 mg), CD8+ cells (200 mg), CD4+ cells (200 mg), or MAC4 to further explore the mechanism by which small numbers of B cIg were administered i.p. day 22, 21, 3, 7, and 11 relative to anti-DR5 cells entering tumors might enable anti-DR5–mediated tumor treatment. In experiments examining the immune cell subsets involved to + the natural suppression of MC38 tumor growth, mice were treated with the apoptosis and the direct or indirect role CD11c DCs might aforementioned neutralizing/depleting Abs 1 d prior to tumor inoculation; contribute to the antitumor activity of Ab-based immunotherapeu- subsequent injections of these mAbs were administered every 4 d through- tics. The findings also create new possibilities in the rational de- out the course of the experiment. Clodronate-containing liposomes for the sign of improved combination cancer therapies comprising the depletion of phagocytic cells and empty-control liposome preparations was used as described [clodrolip; sodium clodronate tetrahydrate, Farchemia death-inducing anti-DR5 mAb. SRL, Treviglio, Italy (20)]. Clodrolip and control liposomes were diluted in PBS and injected i.p. at 2 mg per 20 g body weight, 2 d prior to anti- Materials and Methods DR5 therapy. Subsequent doses of depleting and control liposome preparations (1 mg/20 g body weight) were administered every 2 d. CD11c.DOG Mice mice were treated with 8 ng/g body weight of diphtheria toxin (DT) i.p. on 2 C57BL/6 (B6) and BALB/c mice at 8–12 wks of age were from The Walter day 2, 0, 4, 8, 12, and 16 relative to anti-DR5 treatment. and Eliza Hall Institute of Medical Research. The B6 RAG-1–deficient (RAG-12/2), TRAIL2/2, gld2/2, perforin2/2 (pfp), TNF2/2, and mMT Cytotoxicity assay mice were derived from B6 embryonic stem cells and bred and maintained 51 at the Peter MacCallum Cancer Centre. The B6 and BALB/c DR5-deficient The cytotoxic activity of the MD5-1 mAb was tested in a 20 h [ Cr] release mice (DR52/2) were generated as described (16). The B6 CD11c.DOG assay as previously described (21). Tumor cell susceptibility to TRAIL- mice (17) were kindly provided by Dr. Christian Engwerda (Queensland mediated cytotoxicity was assessed using the murine TRAIL-transfected Institute of Medical Research, Queensland, Australia). All mice were 2PK-3 or mock-transfected 2PK-3 lines as effector cells at the indicated maintained under specific pathogen-free conditions and used in accordance effector/tumor target cell ratios. In the case of MD5-1–mediated with the institutional guidelines of the Peter MacCallum Cancer Centre. cytotoxicity, tumor cells were 1) cultured for 20 h on Protein A plates Animal care was provided in accordance with the procedures outlined in (Pierce, Rockford, IL) coated with 1–10 mg/ml of MD5-1 or UC8-1B9 cIg or 2) cultured in the presence of FcR+ P815 cells, enriched naive splenic the National Institutes of Health Guide for the Care and Use of Laboratory + + Animals. B cells, bone marrow-derived DCs or enriched splenic CD8 or CD4 DCs at the indicated effector/target ratios. B cell enrichment from wild-type Cell lines (WT) B6 mouse spleens was performed as previously described (22). Briefly, collagenase-treated single-cell splenic preparations were fraction- The B6-derived TRAIL-sensitive MC38 colon adenocarcinoma and ated on a cold Percoll gradient of 50, 65, and 80%. Naive B cell enrich- AT-3 mammary carcinoma tumor lines, 2PK-3 mock and mouse TRAIL ment (90–95%) was subsequently achieved using a B cell isolation kit (mTRAIL) transfected B cell lymphoma lines and the P815 mastocytoma (Miltenyi Biotec, Bergisch Gldbach, Germany). Naive B cell purity was line have been described (10, 18, 19). These lines were cultured in DMEM verified by flow cytometric analysis using Ab mixtures containing anti- The Journal of Immunology 3 mouse CD19-FITC (ID3; BD Pharmingen), together with B220-PE (RA3- and MHC I on its surface (Fig. 1A). The MC38 tumor cells were 6B2; eBioscience) mAbs and IgD-FITC (11–26c; eBioscience), together + sensitive to the cytotoxic activity of recombinant mTRAIL, pre- with IgM-PE (II/41; eBioscience) mAbs. Splenic CD11c DC enrichment sented on the surface of the B cell lymphoma line 2PK-3 (Fig. 1B) was performed as described (23). Briefly, WT B6 mouse spleens were collagenase digested at room temperature for 25 mins and single-cell and MD5-1 when cross-linked by FcR on P815 cells in vitro (Fig. suspensions were fractionated on a Nycodenz gradient. To enrich for the 1C). The B6-derived, DR5+ melanoma line, B16-F10 proved re- DCs, a DC isolation kit was used in accordance with the manufacturer’s sistant to the cytotoxic activity of both mTRAIL and MD5-1 (Fig. instructions (Invitrogen). The negatively selected preparations were then 1B,1C). Importantly, we were able to demonstrate the sensitivity of sorted for CD11c+ CD8+ and CD4+ DCs (CD11c-APC [HL3; BD Phar- mingen], CD8-PE [53-6.7; eBioscience], CD4-FITC [RM4-5; eBio- the MC38 tumor line to the tumoricidal actions of MD5-1 in vivo. science]) with a final purify of 90–95%. DC expression of FcRs was Against established MC38 tumors, MD5-1 mAb treatment induced determined by flow cytometric analysis using the 2.4G2 mAb (anti- significant tumor regression, which in some instances resulted in FcgRI/II/III; generated in house), followed by an anti-rat Ig secondary complete tumor rejection (Fig. 2A). Strikingly, the tumor suppres- (goat–anti-rat IgG [H and L chains]-PE; Invitrogen) or an anti- sive effects of MD5-1 were observed as early as 24 h after Ab CD16/32 mAb (CD16/32-Pacific Blue [clone 93; anti-FcgRII/III; eBioscience]). Bone marrow-derived DCs from whole mouse bone marrow administration. The cytotoxic effects of MD5-1 in these estab- were prepared by culturing bone marrow cells harvested from mouse fe- lished MC38 tumors was primarily mediated through DR5 recep- murs in RPMI 1640 supplemented with 10% FCS, 100 mg/ml penicillin/ tor signaling of apoptosis, as the antitumor activity of MD5-1 was streptomycin, 2 mM L-glutamine, 10 mM essential amino acids, 10 mM completely abrogated in mice bearing established MC38 tumors pyruvate, 55 mM 2-ME, and 20 ng/ml GM-CSF. overexpressing cFLIP (Fig. 2A); a protein that can inhibit cas- Therapy of transplanted tumors pase-8–mediated death receptor signaling. Thus, the contribution

of ADCC during the primary rejection phase of anti-DR5 therapy Downloaded from Groups of five mice were injected with 106 MC38, 106 AT-3 or 2 3 105 4T1.2 tumor cells s.c. in the hind flank. Anti-DR5 or UC8-1B9 cIg treat- in established MC38 tumors was minimal. Consistent with this, ment commenced when tumors had developed to a size of 15–20 mm2 we did observe that the antitumor activity of MD5-1 was not (∼6–8 d posttumor inoculation). The administration of depleting or neu- diminished in mice deficient of pfp; a key component of NK tralizing mAbs to some groups of mice or DT to CD11c.DOG mice was cell-mediated ADCC (Fig. 2B); however, we cannot formally initiated 2 d prior to MD5-1 or UC8-1B9 cIg treatment. Tumor size was exclude the possible involvement of other modes of ADCC. A measured every 2–3 d using electronic calipers and data represented as the mean 6 SEM. In some experiments tumor-bearing mice were sacrificed 6– http://www.jimmunol.org/ 8 d posttumor implantation for immune cell characterization. Dissected tumors were cut into small pieces and then digested for 25 mins at 37˚C with 1.6 mg/ml collagenase type IV (Worthington Biochemical, Lake- wood, NJ) and 30 U/ml DNase (Sigma-Aldrich) before being passed through a 70 mm filter. Single-cell suspensions were washed once in media prior to being processed for flow cytometric analysis as described earlier.

TUNEL staining Groups of three to four B6 WT, mMT, and CD11c.DOG mice were in- oculated s.c. with 106 MC38 tumor cells. Mice bearing 20–25 mm2 tumors by guest on September 28, 2021 were treated with MD5-1 or UC8-1B9 cIg 100 mg/mouse i.p. Mice were sacriﬁced 8 and 24 h posttreatment and tumors dissected and processed for immunohistochemical analysis. Fixed and parafﬁn-embedded tissues were randomly sectioned and stained for apoptotic cells using an ApopTag Peroxidase In Situ Apoptosis Detection Kit (Millipore) as per the manufacturer’s instructions. To quantitate the percentage TUNEL positive staining per tumor section the Metomorph Imaging Program Series 7.6.3 was used. Areas of necrosis were excluded from the analysis.

Statistical analysis Statistical signiﬁcance of tumor growth was assessed using a Mann- Whitney U test.

Results Sensitivity of the mouse adenocarcinoma line, MC38 to anti- DR5–mediated apoptosis Previously, we demonstrated that FcR-expressing effector cells were necessary for the anti-DR5 mAb, MD5-1, to induce TRAIL- receptor oligomerization and tumor cell apoptosis in vitro and in vivo (10). To perform a comprehensive analysis of the speciﬁc innate and/or adaptive FcR+ immune cells that may facilitate the FIGURE 1. Characterization of the DR5 expression levels and TRAIL tumoricidal actions of anti-DR5 mAb in vivo, we searched for sensitivity of the B6 derived colon adenocarcinoma line MC38 in vitro. A, a TRAIL-sensitive tumor line derived from the C57BL/6 (B6) Surface expression of DR5 and MHC I on the MC38 and 4T1.2 tumor lines strain, so as to take advantage of a broad range of gene-targeted was assessed by ﬂow cytometry. The open/bold histograms represent DR5 mice available on this background. Of the B6-derived tumor lines or MHC I staining as indicated. The open/dashed histogram represents cIg staining. Solid histograms represent the unstained tumor controls. B and C, that we screened for TRAIL sensitivity, the colon adenocarcinoma The 20 h [51Cr] release assays were used to assess the sensitivity of MC38, tumor line MC38 proved highly sensitive to this death-inducing 4T1.2, and B16-F10 tumor cells to 2PK-3–mTRAIL effectors (B)or signaling pathway. Similar to the BALB/c-derived 4T1.2 tumor MD5-1 cross-linked by FcR+ P815 cells (C). Levels of [51Cr] detected line, which was used for the initial characterization of the antitu- in control wells containing mock-transfected 2PK-3 cells or UC8-1B9 mor activity of the MD5-1 mAb (10), the MC38 tumor line cIg were typically ,5% of maximum (data not shown). Each group was expressed equivalent levels of the TRAIL death receptor DR5 performed in triplicate. Results are representative of three experiments. 4 TUMORICIDAL MECHANISMS OF ANTI-DR5 mAb IN ESTABLISHED CANCER

FIGURE 2. The tumoricidal activity of MD5-1 is dependent on caspase-mediated signaling via tumor cell DR5, but independent of host pfp, TRAIL, FasL, TNF, and DR5 expression. A, To assess the therapeutic efﬁcacy and mechanism of action of the MD5-1 mAb in vivo, 106 MC38 parental tumor cells, or MC38-cFLIP tumor cells were injected s.c. in- to B6 mice. B–F, To identify the death- Downloaded from inducing effector pathways contributing to the tumoricidal activity of MD5-1 in vivo, parental MC38 tumor cells (106) were injected s. c. into B6 WT or pfp 2/2 (B), TRAIL2/2 (C), FasL (gld)2/2 (D), TNF2/2 (E), or DR52/2 (F) mice. A–F, Tumor-bearing mice were treated with MD5-1 or UC8-1B9 cIg at the indicated http://www.jimmunol.org/ time points (designated by arrows). Results are presented as mean tumor growth 6 SEM. B, Statistical differences in tumor sizes between MD5-1 and cIg treated WT [pp # 0.008 (C); p # 0.02 (D); p # 0.005 (E); p # 0.02 (F)] and gene-targeted [ppp # 0.04 (B); p # 0.02 (C), p # 0.009 (D, E); p # 0.04 (F)] mice were determined by the Mann-Whitney U test. Data are representative of two to three experiments. by guest on September 28, 2021 Parentheses indicate the number of surviving mice.

role for the death-inducing ligands, TRAIL, FasL, and TNF were Contribution of NK cells, CD11b+ cells and macrophages to also discounted, as mice deficient of these ligands still demon- the tumoricidal effects of MD5-1 in established s.c. tumors strated MD5-1–mediated suppression of MC38 tumor growth In an earlier study, we proposed that CD11b+ cells, such as mac- (Fig. 2C–E). rophages, were the primary mediators of MD5-1 activity against Having established tumor cell sensitivity to anti-DR5 therapy freshly implanted 4T1.2 tumors cells (10). In this study, we exam- in vivo, we next examined whether host endogenous expression ined whether such cells might facilitate MD5-1–mediated primary of DR5 might impact the therapeutic efficacy of MD5-1 treatment. rejection of established MC38 tumors. Like 4T1.2 tumors (Fig. 3B), In DR5-deficient mice, we found that the antitumor activity of the tumoricidal activity of MD5-1 in established MC38 lesions was MD5-1 was similar to that observed in WT mice, discounting retained following anti-asialoGM1 Ab-mediated depletion of NK any contribution of host DR5 to the antitumor effects of MD5-1 cells (Fig. 3A). In contrast, anti-CD11b mAb treatment, which com- in vivo (Fig. 2F). A similar comparative study in MD5-1–treated pletely abrogated the tumoricidal effects of MD5-1 treatment in 2 2 BALB/c WT and DR5 / mice bearing established s.c. 4T1.2 developing 4T1.2 tumors (Fig. 3D), had no effect on the efficacy lesions confirmed these findings (data not shown). It should be of MD5-1 treatment in the MC38 tumors (Fig. 3C). noted that no signs of mAb-related hepatotoxicity were observed Because flow cytometric analysis of immune cell subsets in these experiments, in which a total of 300 mg MD5-1 was ad- within established MC38 tumors revealed the presence of F4/80+ ministered over three separate treatments. macrophages (19 6 5% of live tumor cells; n =6),wenext The Journal of Immunology 5

FIGURE 3. Contribution of NK cells, CD11b+ cells, and macrophages to the antitumor effects of MD5-1 in established MC38 and 4T1.2 tumors. Mice were injected s.c. with 106 parental MC38 (A, C, E) or 4T1.2 (B, D, F) tumor cells and subsequently treated with Abs to asialoGM1 (A, B) or CD11b (C, D) or the liposome preparation, clodrolip (E, F). Control mice were treated with a matched rat cIg (MAC4) (A–D) or an empty liposome preparation [Empty L; (E)]. Mice were subsequently treated with MD5-1 or UC8-1B9 cIg at the indicated time points (designated by arrows). Results are presented as mean tumor growth 6 SEM. A, C,and E, Statis- tical differences in tumor sizes between the MD5-1

and UC8-1B9 cIg treated WT (pp # 0.008), NK cell- Downloaded from depleted (ppp # 0.02), anti-CD11b treated (ppp # 0.008) or clodrolip treated (pppp # 0.03) mice were determined by the Mann-Whitney U test. E, A statistical difference in tumor sizes was also observed between the UC8-1B9 cIg and UC8-1B9 cIg/clodrolip treated groups (ppp # 0.04). B, D, and F, Statis- tical differences in tumor sizes between MD5-1 and http://www.jimmunol.org/ UC8-1B9 cIg treated WT (pp # 0.008), NK cell- depleted (ppp # 0.008), and clodrolip treated (ppp # 0.03) mice were identiﬁed. F, A statistical loss in MD5-1 activity was also observed in clodrolip-treated mice (pppp # 0.05). Results are representative of two experiments. Parentheses indicate the number of surviving mice. by guest on September 28, 2021

assessed to what extent these cells might contribute to the anti- to contribute minimally to the control of 4T1.2 tumor growth. tumor effects of MD5-1 in established MC38 tumors. For these However, clodrolip treatment did partially inhibit the antitumor experiments, mice bearing established tumors were treated with activity of MD5-1 verifying that macrophages can play a role in clodrolip, a reagent that has proven effective in the selective facilitating the tumoricidal effects of this death inducing mAb in depletion of phagocytic cells, such as macrophages, including some tumor microenvironments. CD11b+ tumor-associated macrophages (24). As a control, some MD5.1 is ineffective in B cell-deficient mice and CD11c+ groups of mice were treated with an empty liposome preparation. DC-deficient mice Treatment of tumor-bearing mice with i.p. injections of clodrolip was found to be sufficient to promote the natural suppression of In the past, the lack of appropriate mAbs or BALB/c gene-targeted established MC38 lesions, because these tumors grew out at mice has restricted our ability to examine the contribution of FcR+ aslowerratethanthoseinmicetreatedwithahamstercIgalone leukocytes, such as B cells and CD11c+ DCs, to MD5-1–mediated (Fig. 3E). These data suggested that macrophages present within tumor suppression. Using gene-targeted mice on a B6-background the established MC38 tumors were supporting tumor progres- we looked to identify whether such adaptive or innate immune sion. Thus, depletion of these cells from the tumor micro- cells could help in facilitating the tumoricidal activity of environment might account for the enhanced therapeutic MD5-1 in established MC38 tumors. Initially, we chose to explore efficacy of MD5-1 when delivered postclodrolip treatment. In- whether MD5-1–mediated primary rejection was dependent on deed, a tumor rejection rate of 50% was recorded in the MD5- T and B cells. Having identified the immunogenic nature of the 1/clodrolip treated group; whereas, no tumor free mice were MC38 tumors and that CD8+ T cells, but not CD4+ T or NK cells, reported for any of the other treatments. It should be noted that have a significant role to play in the natural suppression of these the empty liposome preparation had no effect on the antitumor tumors (Fig. 4A), we first examined whether depletion of CD8+ activity of MD5-1. The maintained efficacy of anti-DR5 therapy cells would limit the tumoricidal activity of MD5-1 in established in clodrolip-treated mice (Fig. 3E) suggested that tumor- s.c. tumors. Although depletion of CD8+ T cells significantly en- associated macrophages were not the primary mediators of the hanced MC38 tumor growth as compared with cIg-treated mice cytotoxic effects of MD5-1 in established MC38 tumors. Clo- (Fig. 4B), MD5-1 treatment still inhibited tumor growth in a sim- drolip was also administered to mice bearing established 4T1.2 ilar manner to that observed in WT mice with an intact T cell tumors (Fig. 3F). In this tumor setting, macrophages were found compartment (Fig. 4B). These data suggested that CD8+ T cells 6 TUMORICIDAL MECHANISMS OF ANTI-DR5 mAb IN ESTABLISHED CANCER

(26). However, these findings were intriguing and somewhat coun- terintuitive to the low frequency of B cells in established MC38 tumors (0.05 6 0.02% total live CD19+B220+ cells; n = 6). A possible role for DCs in MD5-1 antitumor activity in vivo was speculated based on their infiltration into 4T1.2 tumors post Ab- treatment (10); however, confirmation of their involvement has not been possible in the BALB/c strain. Certainly, DCs do express both activating and inhibitory FcgRs and thus have the potential to facilitate MD5-1–mediated cytotoxicity (27). For this we employed the use of CD11c.DOG transgenic mice, which can largely be depleted (.90%) of CD11chigh MHC class II+ conventional DCs and 50–60% of plasmacytoid DCs by continuous injections of DT [(17) and communication with Prof. G. Ha¨mmerling]. In these experiments, CD11c.DOG mice bearing established MC38 tumors were treated with DT 2 d prior to MD5-1 or hamster cIg treatment. Strikingly in the DT-treated CD11c.DOG mice, which possess an intact B cell compartment, the tumoricidal activity of MD5-1 was almost completely abrogated, with the tumors grow- Downloaded from ing out in all mice in a similar manner to that seen in cIg-treated groups (Fig. 4D). These data identified an important role for + FIGURE 4. The tumoricidal activity of MD5-1 therapy against immuno- CD11c DCs in MD5-1–mediated antitumor activity. genic MC38 tumors is retained in mice depleted of CD8+ T cells but lost in Importantly, we were able to further validate the importance B cell-deficient and CD11c+ DC-deficient mice. A, B6 WT mice were of an intact B cell compartment to the antitumor activity of MD5-1 injected i.p. with mAbs to CD4, CD8, or asialoGM-1. Control groups of using the B6-derived mammary carcinoma line AT-3, which like the 6 http://www.jimmunol.org/ mice were treated with PBS or MAC4 cIg. One-day post-Ab treatment 10 MC38 line, expresses both DR5 and MHC I (Fig. 5A) and is sensitive MC38 tumor cells were injected s.c in the hind flank. Results are presented to anti-DR5–mediated cytotoxicity in vitro (Fig. 5B). In both mMT as mean tumor growth 6 SEM. Statistical differences in tumor sizes between MAC4 cIg and anti-CD8 treated mice were determined by the Mann- Whitney U test (pp # 0.008). B, Mice were injected s.c. with 106 MC38 tumor cells and treated with an anti-CD8a mAb or MAC4 cIg on day 6 posttumor inoculation. Mice were subsequently treated with MD5-1 or UC8- 1B9 cIg at the indicated time points (designated by arrows). Results are presented, as mean tumor growth 6 SEM. Statistical differences in mean p tumor size between the UC8-1B9 cIg and MD5-1–treated WT ( p # 0.02) by guest on September 28, 2021 and CD8-depleted (pp # 0.02) mice were detected. C,WT,RAG-12/2,and mMT mice were injected s.c. with 106 MC38 parental tumor cells. On day 6 after tumor inoculation mice were injected with MD5-1 or UC8-1B9 cIg at the designated time points (indicated by arrows). Results are presented as mean tumor growth 6 SEM. Statistical differences in mean tumor size between the UC8-1B9 cIg and MD5-1–treated WT mice were detected (pp # 0.008). D, CD11c.DOG mice were injected s.c. with 106 MC38 tumor cells. Tumor-bearing mice were treated with DT or PBS prior to receiving MD5-1 or UC8-1B9 cIg at the designated time points (designated by arrows). Results are presented as mean tumor growth 6 SEM. All in vivo results are representative of two experiments. were not critical for the tumor growth inhibitory effects of MD5-1. Surprisingly, however, in RAG-1–deficient mice, the tumoricidal effects of MD5-1 were completely abrogated, reaffirming the likely contribution of adaptive immune components to MC38 FIGURE 5. The tumoricidal activity of MD5-1 against AT-3 mammary tumor cells is dependent on adaptive immune components. A, Surface growth suppression post–anti-DR5 therapy (Fig. 4C). Again, this expression of DR5 and MHC I on the AT-3 tumor line was assessed by finding was in contrast to what we observed in the 4T1.2 tumors flow cytometric analysis. The open/bold histograms represent DR5 or where MD5-1 treatment was equally effective in WT and SCID MHC I staining as indicated. The open/dashed histogram represents cIg mice (data not shown). staining and the solid histograms represent the unstained tumor controls. B, Given that B cells have been reported to express the inhibitory The comparative sensitivity of the MC38 and AT-3 tumor cells to MD5-1– low-affinity FcR for IgG, FcgRIIB, which regulates activating mediated apoptosis was assessed in a 20 h [51Cr] release assay, in which signals transduced by the B cell receptor (25), we next examined the tumor cells were cocultured with FcR+ P815 cells and MD5-1 or UC8- the tumoricidal effects of MD5-1 in B cell-deficient, mMT mice. 1B9 cIg. Each group was performed in triplicate. Results are representative of three experiments. C and D,WT,mMT, or RAG-12/2 mice were Strikingly in these mice, MD5-1 treatment had no therapeutic 6 benefit and the MC38 tumors grew out in a similar manner to that injected s.c. with 10 AT-3 tumor cells. On day 8 after tumor inoculation mice were injected with MD5-1 or UC8-1B9 cIg at the indicated time observed for the cIg-treated tumors (Fig. 4C). It is unlikely that points (designated with arrows). Results are presented as mean tumor the observed loss of MD5-1 antitumor activity was due to altered growth 6 SEM. Statistical differences in tumor sizes between MD5-1– pharmacokinetics of the MD5-1 mAb in mMT mice (lacking Ig), and cIg-treated mice were determined by the Mann-Whitney U test (pp # because similar doses of an anti-mouse CD137 (4-1BB) mAb was 0.008). No statistical differences between the cIg- and MD5-1–treated reported to be comparatively effective in both WT and mMT mice mMT and Rag12/2 mice were detected. The Journal of Immunology 7

(Fig. 5C) and RAG-12/2 (Fig. 5D) mice, MD5-1 treatment failed CD8+ and CD4+ DCs we chose to compare the capacity of these to mount a response against established s.c. AT-3 tumors. two DC subsets to facilitate FcR–dependent MD5-1–mediated cytotoxicity against the MC38 tumor line in vitro (Fig. 6B). As B cells and CD11c+ DC are capable of triggering MD5-1– expected, tumor cell death was dictated by the level of FcR mediated apoptosis in vitro expression on the DC subsets. Consistent with the need to cross- Based on these in vivo observations, we were interested in de- link the MD5-1 mAb to signal apoptosis, Ab-mediated blockade of + termining whether B cells and CD11c DC could directly trigger FcRs on both DC subsets effectively inhibited the cytotoxic activ- MD5-1–mediated cytotoxicity in vitro. Indeed, we detected FcR ity of this death-inducing mAb. BMDCs were also observed to be + + on ex vivo splenic CD8 , CD4 , and plasmacytoid DCs (Fig. 6A). capable of mediating MD5-1 cytotoxicity against the MC38 tumor Interestingly, of these DC subsets, expression of the activating and line (data not shown). Although in vivo, it is unlikely that CD8+ inhibitory FcgRII/FcgRIII receptors (CD16/32) was only detected DCs contribute to the antitumor effects of MD5-1, because Ab- + on CD8 DCs, the primary mediators of Ag cross-presentation mediated depletion of CD8a+ cells did not impair the ability of (Fig. 6A) (28). Given the differential expression of FcRs on MD5-1 to suppress tumor growth (Fig. 4B), it remained possible that other CD11c+ DC subsets were important (Fig. 6A). We were also able to demonstrate that enriched splenic B cells could facilitate MD5-1–mediated apoptosis of the MC38 tumor cells and this was completely abrogated by Ab-mediated blockade of FcR on the B cells (Fig. 6C). Overexpression of cFLIP in the MC38 tumor

line was also shown to block the cytotoxic effects of MD5-1 in this Downloaded from in vitro system (Fig. 4C). These results conﬁrmed that B cell- mediated cross-linking of the MD5-1 mAb was sufﬁcient to induce DR5- and caspase-mediated cell death. Only B cells are critical for MD5-1–mediated tumor cell apoptosis in vivo http://www.jimmunol.org/ To determine whether B cells and CD11c+ DCs were directly involved in MD5-1–mediated apoptosis in vivo, we used a TUNEL staining assay to quantitatively compare the levels of MD5-1–mediated tumor cell apoptosis in solid MC38 tumors isolated from WT-, mMT-, and DT-treated CD11c.DOG mice, 8 and 24 h post-Ab treatment (Fig. 7; data not shown). Any observed loss in TUNEL staining within tumor sections from B cell by guest on September 28, 2021

FIGURE 6. B cells and CD11chi DC subsets are capable of MD5-1– mediated cytotoxicity in vitro. A, Surface expression of FcgRs on splenic CD8+, CD4+, and pDC subsets was assessed by flow cytometry. FcgRI/II/ FIGURE 7. Detection of MD5-1–mediated apoptosis in established III expression was detected with the 2.4G2 mAb (top row, open/bold MC38 tumors. WT, mMT, or CD11c.DOG mice were injected s.c. with histogram). A CD16/32 mAb was used to detect expression of the 106 MC38 tumor cells. Mice bearing established MC38 tumors (20–25 inhibitory FcRs, FcgRIII/II (bottom row, open/bold histograms). The mm2) were treated with MD5-1 or UC8-1B9 cIg. CD11c.DOG mice were open/dashed histograms (bottom row) demonstrate the level of CD16/32 treated with DT days 22 and 0 relative to anti-DR5 treatment. Mice were staining posttreatment of splenic DCs with the control 2.4G2 FcR-blocking sacrificed at 8 and 24 (data not shown) h post-mAb treatment and tumors mAb. The solid histograms represent unstained DCs. B, Cytotoxic activity harvested and processed for TUNEL staining. Whole tumor sections were of MD5-1 mAb against MC38 tumor cells cultured with enriched ex vivo quantitatively analyzed for percentage TUNEL positive staining per tumor splenic CD8+ or CD4+ DCs at the indicated effector/target ratios. Control section area using the Metomorph Imaging Program Series 7.6.3. The cultures contained UC8-1B9 cIg or DCs preincubated with the FcR- closed circles represent the average percentage area of TUNEL staining blocking Ab 2.4G2. Each group was performed in triplicate. C, Cytotoxic detected in three random tumor sections from an individual mouse/tumor. activity of MD5-1 against MC38 and MC38-cFLIP tumor lines when The histogram represents the average percentage of TUNEL staining for cultured in the presence of enriched naive splenic B cells at a 20:1 and each group of mice. Tumors of a similar size (area as determined in pixels) 5:1 effector/target ratio. In some groups, B cells were treated with an FcR- were used for this analysis. Statistical differences in percentage TUNEL blocking Ab (2.4G2) prior to being cultured with MC38 cells and MD5-1. staining between MD5-1– and cIg-treated tumors from WT and CD11c. Control cultures contained UC8-1B9 cIg or B cells alone (no Ab present). DOG mice (pp # 0.0002) and MD5-1–treated tumors from WT and mMT Each group was performed in triplicate. Results are representative of three mice (ppp # 0.0001) were determined by the Mann-Whitney U test using experiments. the % TUNEL staining values from all tumor sections analyzed per group. 8 TUMORICIDAL MECHANISMS OF ANTI-DR5 mAb IN ESTABLISHED CANCER and/or CD11c+ cell-deficient mice, compared with that observed present in large numbers within the 4T1.2 tumors (15 6 8% of in tumors isolated from MD5-1–treated WT mice, would identify live tumor cells, n = 3). Certainly, neutrophils do posses potent whether one or both of these immune cell subsets was key in cytotoxic activity that can be enhanced in the presence of antitu- MD5-1–mediated apoptosis in vivo. Interestingly, quantitative mor mAbs (31). Interestingly, in the MC38 tumor model we iden- analysis of TUNEL positive staining in random whole tumor tified B cells as critical for MD5-1–mediated apoptosis and sections from MD5-1– and cIg-treated tumors, 8 h posttreatment, antitumor activity. These data were surprising given the low revealed a significant loss of MD5-1–induced TUNEL activity frequency of B cells detectable within established MC38 tumors. when B cells, but not CD11c+ DCs, were absent (Fig. 7). However, the ex vivo analysis of a tumor specimen only provides Similar results were also evident at 24 h post-Ab treatment; how- a momentary snapshot of what is occurring within that tumor, and ever, consistent with a rapid apoptotic effect, the level of detect- thus we cannot discount that the circulation of B cells into and out able apoptosis was considerably reduced at that later time point of the tumor microenvironment may be sufficient to enable MD5- (data not shown). Ultimately, these data identify B cells as being 1–bound B cells to trigger DR5-signaling of tumor cell apoptosis. directly involved in, or upstream of, the first primary event in What factors may promote this preferential interaction between tumor suppression—MD5-1–mediated apoptosis in vivo. DT- B cells and MD5-1 within the tumor microenvironment and/or treated CD11c.DOG mice that possess a fully competent B cell post-mAb administration is currently under investigation. compartment, displayed the same quantity of TUNEL positive Characterization of DC involvement in MD5-1–mediated MC38 tumor cells as WT mice, post–MD5-1 treatment. Impor- antitumor activity was performed using the CD11c.DOG transgenic tantly, these data rule out a direct role for CD11c+ DCs in trig- mouse strain, that when treated with DT can be depleted of .90% of gering MD5-1–mediated apoptosis in vivo, however, suggest that CD11chigh MHC class II+ DCs (17). In these mice, which possess an Downloaded from the profound loss of MD5-1 activity in these mice is a result of intact B cell compartment, MD5-1–mediated tumor cell apoptosis a downstream role for DCs in promoting and/or sustaining the was still detectable; however, surprisingly this alone did not result in antitumor effects of MD5-1 therapy in vivo. the therapeutic control of tumor growth. Ultimately, this result both eliminates a direct role for CD11c+ DCs in facilitating MD5-1– Discussion mediated apoptosis and implicates an important role of DCs

Based on the cell death and immune stimulating properties of downstream of tumor cell apoptosis mediated by MD5-1. Although http://www.jimmunol.org/ mAbs, these targeted therapeutics have the capacity to significantly DT treatment of these mice has also been reported to deplete a small alter the immunogenicity of the tumor microenvironment and in proportion of F4/80+, MOMA-1+, and ERTR-9+ subpopulations of turn increase tumor cell sensitivity to immune attack. Certainly, macrophages in the spleen (17), this is unlikely to have accounted mAbs that kill tumor cells, like anti-DR5, have proven to be an for the loss of MD5-1 efficacy, because clodrolip, which can also effective means of priming tumor-specific cell-mediated immunity target these macrophage populations (24), had no impact on the (10). Yet little is known about how this immunity is primed or capacity of MD5-1 to inhibit the growth of established MC38 even the innate and adaptive cellular components responsible for tumors. The mechanism by which DCs contribute to the antitumor initiating the tumoricidal actions of anti-DR5 mAb therapy. Such activity of MD5-1 in vivo now remains to be resolved. information would ensure that the full therapeutic power of death In the absence of equivalent strains on the BALB/c background, by guest on September 28, 2021 inducing mAbs is best capitalized on in future combination strat- it remains to be tested whether DCs and B cells can also contribute egies. Herein we have identified FcR+ immune cell subsets that are to MD5-1 suppression of 4T1.2 tumors. However, it should be responsible for MD5-1–mediated tumor cell apoptosis and noted that in FcgRIIB-deficient BALB/c mice, MD5-1 exerted MD5-1 antitumor activity in vivo. We report for the first time that an antitumor effect equivalent to WT mice, possibly discounting B cells and CD11c+ DCs can play a critical role in MD5-1– a direct role for B cells in this tumor model (10). Certainly, this mediated primary suppression of established MC38 tumors; previously unreported contribution of B cells and DCs to the anti- whereas, CD11b+ and NK cells were not critical. Importantly, tumor effects of MD5-1 is important and may have far-reaching the contribution of B cells to the antitumor efficacy of MD5-1 clinical implications, particularly if this type of death induction can was also validated in the B6-derived mammary carcinoma model, also promote DC uptake of tumor Ags and boost tumor immunity. AT-3. Both B cells and DCs were identified as being capable of This capacity of death-inducing mAbs like MD5-1 to potentially binding MD5-1 via their FcRs and permitting DR5-mediated ap- bridge dying tumor cells with immune cells that can promote tu- optosis of tumor cells in vitro. These findings contrasted some- mor-specific immunity may have a number of important immuno- what with our previous and present analysis of 4T1.2 tumors in logical consequences. Indeed, although Fc-mediated cross-linking which CD11b+ cells and, to a lesser extent, NK cells were impor- of the MD5-1 mAb is critical to stimulate tumor cell apoptosis, tant in the antitumor activity of MD5-1 (10). Interestingly, we this interaction may in turn help target tumor Ags to APCs and identified that host DR5 expression, which is tightly regulated subsequently allow the development of tumor-specific immunity. on immature DCs (29) and naive and activated B cells (30), did If this is occurring, and CD82 CD11c+ DCs are important, it still not influence the antitumor activity of MD5-1 in vivo. We have remains unclear what final effector mechanism causes tumor sup- further shown that B cells, but not CD11c+ DCs, were critical for pression. Ultimately the, optimization of the targeting of some MD5-1–mediated tumor cell apoptosis in vivo, thus suggesting antitumor mAbs to selective FcgRs on DCs and possibly even that CD11c+ DCs play an important role in the antitumor B cells will in some cases be beneficial to induce an efficient activity of MD5-1 downstream of primary tumor cell apoptosis. and durable effect on the tumor. Despite the distinct tumor microenvironments of TRAIL- It is now becoming increasingly evident that many distinct types sensitive MC38 and 4T1.2 tumors, we found that anti-DR5 therapy of apoptosis-inducing agents have the capacity to render tumor cell was still capable of inducing significant growth inhibition in both death immunogenic and thus prime cellular immune responses that types of established tumors. Previously, we concluded that the an- can be mobilized to eliminate cancer (32, 33). Currently, little is titumor effects of MD5-1 against implanted 4T1.2 tumor cells was known about the immunogenicity of TRAIL-mediated killing; primarily mediated by FcR on CD11b+ macrophages (10). This however, one study has presented evidence to suggest that high- was also observed in established 4T1.2 tumors; however, we could mobility group box 1 protein release from tumor cells, which not discount a possible role for FcR on neutrophils, which are through binding of TLR4 on DCs can stimulate the processing The Journal of Immunology 9 and presentation of Ags to cytotoxic T lymphocytes (34), can be antibody-dependent cellular cytotoxicity against LFA-3 and HER2/neu. Exp. Hematol. 27: 1533–1541. promoted by soluble TRAIL and possibly even by activated NK 4. Albanell, J., J. Codony, A. Rovira, B. Mellado, and P. Gascoń. 2003. Mechanism cells that express TRAIL (35). Given that we did not identify a role of action of anti-HER2 monoclonal antibodies: scientific update on trastuzumab for CD8+ T cells in the antitumor responses mediated by MD5- and 2C4. Adv. Exp. Med. Biol. 532: 253–268. 5. Uchida, J., Y. Hamaguchi, J. A. Oliver, J. V. Ravetch, J. C. Poe, K. M. Haas, and 1, it could be argued that Ab-mediated DR5 signaling of tumor T. F. Tedder. 2004. The innate mononuclear phagocyte network depletes B cell apoptosis is not a strong inducer of immunogenicity. Cer- lymphocytes through Fc receptor-dependent mechanisms during anti-CD20 an- tainly, it will be important to consider whether targeting the tibody immunotherapy. J. Exp. Med. 199: 1659–1669. 6. Takeda, K., K. Okumura, and M. J. Smyth. 2007. Combination antibody-based TRAIL death receptors, DR5 and DR4 alone or in combination cancer immunotherapy. Cancer Sci. 98: 1297–1302. will have any bearing on the immunogenic outcomes of signaling 7. Mahalingam, D., E. Szegezdi, M. Keane, S. Jong, and A. Samali. 2009. TRAIL TRAIL-mediated apoptosis. However, although killing tumor cells receptor signalling and modulation: Are we on the right TRAIL? Cancer Treat. Rev. 35: 280–288. in a manner that generates danger signals and tumor Ags in the 8. Sheridan, J. P., S. A. Marsters, R. M. Pitti, A. Gurney, M. Skubatch, D. Baldwin, right context can promote engagement of innate and adaptive L. Ramakrishnan, C. L. Gray, K. Baker, W. I. Wood, et al. 1997. Control of TRAIL-induced apoptosis by a family of signaling and decoy receptors. Science immunity, this response will likely prove insufficient in controlling 277: 818–821. established cancer because of the immunosuppressive nature of 9. Kelley, S. K., L. A. Harris, D. Xie, L. Deforge, K. Totpal, J. Bussiere, and the tumor microenvironment. Indeed, we have evidence demon- J. A. Fox. 2001. Preclinical studies to predict the disposition of Apo2L/tumor necrosis factor-related apoptosis-inducing ligand in humans: char- strating that MC38 tumors are sensitive to T regulatory cell con- acterization of in vivo efficacy, pharmacokinetics, and safety. J. Pharmacol. Exp. trol and express high levels of the immune inhibitory program Ther. 299: 31–38. death-1 receptor ligand PD-L1 (data not shown), that may collec- 10. Takeda, K., N. Yamaguchi, H. Akiba, Y. Kojima, Y. Hayakawa, J. E. Tanner, T. J. Sayers, N. Seki, K. Okumura, H. Yagita, and M. J. Smyth. 2004. Induction tively maintain a tolerogenic environment within the establishing of tumor-specific T cell immunity by anti-DR5 antibody therapy. J. Exp. Med. Downloaded from tumors. We should also be mindful of the fact that Abs or immune 199: 437–448. complexes can trigger activating and inhibitory signaling path- 11. Lowe, S. W., S. Bodis, A. McClatchey, L. Remington, H. E. Ruley, D. E. Fisher, D. E. Housman, and T. Jacks. 1994. p53 status and the efficacy of cancer therapy ways and depending on the individual strength of these opposing in vivo. Science 266: 807–810. signals, will determine whether DCs become stimulatory versus 12. Cretney, E., K. Takeda, H. Yagita, M. Glaccum, J. J. Peschon, and M. J. Smyth. tolerogenic (27). 2002. Increased susceptibility to tumor initiation and metastasis in TNF-related apoptosis-inducing ligand-deficient mice. J. Immunol. 168: 1356–1361.

Notably, the capacity of mAbs to engage and/or stimulate FcRs 13. Smyth, M. J., E. Cretney, K. Takeda, R. H. Wiltrout, L. M. Sedger, N. Kayagaki, http://www.jimmunol.org/ will likely be dictated by their isotype. Indeed, the antitumor ac- H. Yagita, and K. Okumura. 2001. Tumor necrosis factor-related apoptosis- inducing ligand (TRAIL) contributes to interferon gamma-dependent natural tivity of antihuman TRAIL-R1 mAbs in mice bearing human tumor killer cell protection from tumor metastasis. J. Exp. Med. 193: 661–670. xenografts was found to be markedly influenced by their isotype, 14. Takeda, K., M. J. Smyth, E. Cretney, Y. Hayakawa, N. Kayagaki, H. Yagita, and suggesting a critical contribution of FcRs for some effector func- K. Okumura. 2002. Critical role for tumor necrosis factor-related apoptosis- inducing ligand in immune surveillance against tumor development. J. Exp. tions in vivo (36). Similar observations were reported for anti- Med. 195: 161–169. CD20 targeting mAbs (5). However, due to the hamster origin 15. Oldenhuis,C.N.,J.H.Stegehuis,A.M.Walenkamp,S.deJong,andE.G.deVries. of the MD5-1 mAb, its isotype remains undefined and thus such 2008. Targeting TRAIL death receptors. Curr. Opin. Pharmacol. 8: 433–439. 16. Diehl, G. E., H. H. Yue, K. Hsieh, A. A. Kuang, M. Ho, L. A. Morici, L. L. Lenz, studies have not been pursued. It is noteworthy though that not all D. Cado, L. W. Riley, and A. Winoto. 2004. TRAIL-R as a negative regulator of anti-TRAIL targeting mAbs that are being tested for preclinical innate immune cell responses. Immunity 21: 877–889. by guest on September 28, 2021 development or in clinical trials need to be cross-linked to trigger 17. Hochweller, K., J. Striegler, G. J. Ha¨mmerling, and N. Garbi. 2008. A novel CD11c.DTR transgenic mouse for depletion of dendritic cells reveals their re- tumor cell apoptosis; a phenomenon likely dictated by the affinity quirement for homeostatic proliferation of natural killer cells. Eur. J. Immunol. of the Fab9 binding domain for its target Ag. However, regardless 38: 2776–2783. 18. Frew, A. J., R. K. Lindemann, B. P. Martin, C. J. Clarke, J. Sharkey, of whether an Ab needs to be cross-linked or not, optimization of D. A. Anthony, K. M. Banks, N. M. Haynes, P. Gangatirkar, K. Stanley, et al. Ab design so as to enhance its capacity to interact with stimulatory 2008. Combination therapy of established cancer using a histone deacetylase FcgRs on DCs and/or phagocytic cells will be critical to promote inhibitor and a TRAIL receptor agonist. Proc. Natl. Acad. Sci. USA 105: 11317– 11322. their function in vivo. 19. Stewart, T. J., D. J. Liewehr, S. M. Steinberg, K. M. Greeneltch, and Ultimately, we predict that in combination with other immune S. I. Abrams. 2009. Modulating the expression of IFN regulatory factor 8 alters stimulatory agents, the antitumor activity of anti-DR5 or -DR4 ther- the protumorigenic behavior of CD11b+Gr-1+ myeloid cells. J. Immunol. 183: 117–128. apy will greatly improve. Importantly in this study, we have iden- 20. Seiler, P., P. Aichele, B. Odermatt, H. Hengartner, R. M. Zinkernagel, and tified tumor-associated immunological targets that we can now R. A. Schwendener. 1997. Crucial role of marginal zone macrophages and attempt to manipulate using both conventional and experimental marginal zone metallophils in the clearance of lymphocytic choriomeningitis virus infection. Eur. J. Immunol. 27: 2626–2633. reagents to promote their function and further augment the thera- 21. Haynes, N. M., M. J. Smyth, M. H. Kershaw, J. A. Trapani, and P. K. Darcy. peutic efficacy of Ab-based death-inducing immunotherapeutics, 1999. Fas-ligand-mediated lysis of erbB-2-expressing tumour cells by redirected cytotoxic T lymphocytes. Cancer Immunol. Immunother. 47: 278–286. such as anti-DR5 against established disease. 22. Hasbold, J., L. M. Corcoran, D. M. Tarlinton, S. G. Tangye, and P. D. Hodgkin. 2004. Evidence from the generation of immunoglobulin G-secreting cells that Acknowledgments stochastic mechanisms regulate lymphocyte differentiation. Nat. Immunol. 5: 55–63. We thank Michelle Stirling for technical assistance and Fiona Amante from 23. Naik, S. H., D. Metcalf, A. van Nieuwenhuijze, I. Wicks, L. Wu, M. O’Keeffe, Queensland Institute of Medical Research, Queensland, Australia, for help and K. Shortman. 2006. Intrasplenic steady-state dendritic cell precursors that with the screening and shipping of CD11c.DOG mice. are distinct from monocytes. Nat. Immunol. 7: 663–671. 24. Zeisberger, S. M., B. Odermatt, C. Marty, A. H. Zehnder-Fja¨llman, K. Ballmer- Hofer, and R. A. Schwendener. 2006. Clodronate-liposome-mediated depletion Disclosures of tumour-associated macrophages: a new and highly effective antiangiogenic The authors have no financial conflicts of interest. therapy approach. Br. J. Cancer 95: 272–281. 25. Ravetch, J. V., and L. L. Lanier. 2000. Immune inhibitory receptors. Science 290: 84–89. 26. Westwood, J. A., N. M. Haynes, J. Sharkey, N. McLaughlin, H. J. Pegram, References R. A. Schwendener, M. J. Smyth, P. K. Darcy, and M. H. Kershaw. 2009. Toll- 1. Murphy, C. G., and S. Modi. 2009. HER2 breast cancer therapies: a review. like receptor triggering and T-cell costimulation induce potent antitumor immu- Biologics 3: 289–301. nity in mice. Clin. Cancer Res. 15: 7624–7633. 2. Hauptrock, B., and G. Hess. 2008. Rituximab in the treatment of non-Hodgkin’s 27. Nimmerjahn, F., and J. V. Ravetch. 2007. Fc-receptors as regulators of immunity. lymphoma. Biologics 2: 619–633. Adv. Immunol. 96: 179–204. 3. Cooley, S., L. J. Burns, T. Repka, and J. S. Miller. 1999. Natural killer cell 28. Hildner, K., B. T. Edelson, W. E. Purtha, M. Diamond, H. Matsushita, cytotoxicity of breast cancer targets is enhanced by two distinct mechanisms of M. Kohyama, B. Calderon, B. U. Schraml, E. R. Unanue, M. S. Diamond, et al. 10 TUMORICIDAL MECHANISMS OF ANTI-DR5 mAb IN ESTABLISHED CANCER

2008. Batf3 deﬁciency reveals a critical role for CD8alpha+ dendritic cells in 33. Haynes, N. M., R. G. van der Most, R. A. Lake, and M. J. Smyth. 2008. Im- cytotoxic T cell immunity. Science 322: 1097–1100. munogenic anti-cancer chemotherapy as an emerging concept. Curr. Opin. 29. Ursini-Siegel, J., W. Zhang, A. Altmeyer, E. N. Hatada, R. K. Do, H. Yagita, and Immunol. 20: 545–557. S. Chen-Kiang. 2002. TRAIL/Apo-2 ligand induces primary plasma cell apo- 34. Apetoh, L., F. Ghiringhelli, A. Tesniere, M. Obeid, C. Ortiz, A. Criollo, ptosis. J. Immunol. 169: 5505–5513. G. Mignot, M. C. Maiuri, E. Ullrich, P. Saulnier, et al. 2007. Toll-like receptor 4- 30. Leverkus, M., H. Walczak, A. McLellan, H. W. Fries, G. Terbeck, E. B. Bro¨cker, dependent contribution of the immune system to anticancer chemotherapy and and E. Ka¨mpgen. 2000. Maturation of dendritic cells leads to up-regulation of radiotherapy. Nat. Med. 13: 1050–1059. cellular FLICE-inhibitory protein and concomitant down-regulation of death 35. Ito, N., R. A. DeMarco, R. B. Mailliard, J. Han, H. Rabinowich, P. Kalinski, ligand-mediated apoptosis. Blood 96: 2628–2631. D. B. Stolz, H. J. Zeh, III, and M. T. Lotze. 2007. Cytolytic cells induce HMGB1 31. van Egmond, M. 2008. Neutrophils in antibody-based immunotherapy of cancer. release from melanoma cell lines. J. Leukoc. Biol. 81: 75–83. Expert Opin. Biol. Ther. 8: 83–94. 36. Chuntharapai, A., K. Dodge, K. Grimmer, K. Schroeder, S. A. Marsters, H. Koeppen, 32. Zitvogel, L., L. Apetoh, F. Ghiringhelli, and G. Kroemer. 2008. Immunological A. Ashkenazi, and K. J. Kim. 2001. Isotype-dependent inhibition of tumor growth aspects of cancer chemotherapy. Nat. Rev. Immunol. 8: 59–73. in vivo by monoclonal antibodies to death receptor 4. J. Immunol. 166: 4891–4898. Downloaded from http://www.jimmunol.org/ by guest on September 28, 2021