In situ Stimulation of CD40 and Toll-like Receptor 3 Transforms Ovarian Cancer–Infiltrating Dendritic Cells from Immunosuppressive to Immunostimulatory Cells

Dendritic cells are best known for initiating T-cell–mediated immune responses. Consequently, multiple strategies aimed at using the immunostimulatory potential of adoptively administered dendritic cells have been attempted to boost therapeutically relevant immune responses against established human cancers. Unfortunately, although measurable immunologic enhancement has been routinely attained and new promising strategies are under investigation, the emerging view is that the dendritic cell–based vaccination approaches tested thus far do not effectively induce cancer regression in most patients (1, 2).

In the United States alone, epithelial ovarian cancer claimed the lives of >15,000 women in 2008 (3). Regrettably, chemotherapies implemented in the last 30 years have led to a 5-year survival rate of 30%, at best, for patients with metastatic ovarian carcinoma, the stage at which most cases are diagnosed (3). Although the lethality of the disease has traditionally reduced the number of potential advocates, the need for new complementary treatments is increasingly clear. Interestingly, ovarian cancer is an ideal target for novel immunotherapies for several reasons. Firstly, studies pioneered by Coukos and colleagues indicate that ovarian cancer naturally triggers anticancer immune responses (4–7). Secondly, although ovarian cancer is a devastating disease, metastases are frequently restricted to the peritoneal cavity where the tumor microenvironment is directly accessible, which prevents the need for systemic delivery of immunostimulatory treatments. Thirdly, CD11c⁺DEC-205⁺MHC-II^lowCD11b^- dendritic cells expressing low to undetectable levels of costimulatory CD80 represent the most frequent leukocytic subset in the microenvironment of mouse and human solid ovarian carcinomas (8–11), hence promoting their inherent immunostimulatory potential may prove the most effective therapy yet.

To boost endogenous immunity against different epithelial tumors, signaling through CD40 has also been attempted with promising results, although therapeutic effectiveness was limited by toxicity (12, 13). In independent trials, Toll-like receptor (TLR) agonists have also been individually implemented as adjuvants (14). Furthermore, recent reports in nontumor systems indicate that a combinatorial stimulation of TLRs and TRAF signaling by CD40 cross-linking generates a 10- to 20-fold increase in the number of activated CD8⁺ T cells compared with either agonist alone (15). In murine nonepithelial tumors, CD40/TLR7 agonists have been used as adjuvants in vaccination using exogenous tumor antigen, which resulted in stronger and less toxic antitumor memory T-cell responses compared with monotherapy (16).

We have shown previously that CD11c⁺ dendritic cells sorted from ovarian cancer–bearing mice are competent phagocytes but cannot efficiently present ovalbumin to transgenic T cells before receiving an aggressive stimulatory cocktail (9). In addition, the elimination of dendritic cells from tumor locations delays ovarian cancer progression by boosting measurable T-cell–mediated antitumor immunity (17), and previous reports show that human ovarian cancer–associated dendritic cells express functional levels of immunosuppressive PD-L1 (18). However, there is no direct evidence that CD11c⁺MHC-II⁺ dendritic cells from ovarian carcinoma specimens suppress antigen-specific T-cell responses. We hypothesized that, if ovarian dendritic cells acted as bona fide immunosuppressive cells in the ovarian carcinoma microenvironment where they massively accumulate (9, 10), reversing their tolerogenic phenotype in vivo while promoting their capacity to present tumor antigens that they spontaneously engulf could elicit therapeutically effective antitumor immunity. To test this hypothesis, we administered a cocktail of CD40 and TLR3 agonists (16). Preceding synergistic cancer interventions were limited to TLR7 targeting; however, plasmacytoid dendritic cells are the main subset of dendritic cells which express TLR7 (and TLR3 at almost undetectable levels) in both mouse and human (19, 20). We show here that the activation of TLR3, expressed on ovarian tumor–associated CD11c⁺MHC-II⁺ dendritic cells, synergizes with CD40 agonist to induce dendritic cells maturation in situ, which reverses their immunosuppressive phenotype and induces T-cell responses that lead to the rejection of otherwise lethal ovarian carcinomas.

Mice were procured from the National Cancer Institute or The Jackson Laboratory. Experiments were approved by our Institutional Animal Care and Use Committee. Stage III to IV human ovarian carcinoma specimens were procured through Research Pathology Services at Dartmouth under an approved protocol. Single-cell suspensions were generated as described previously (21).

For treatments, wild-type C57BL/6 mice were i.p. injected with 1.5 × 10⁶ ID8-luciferase (17), ID8-GFP, or ID8-Defb29/Vegf-A ovarian carcinoma cells. At days 7, 14, 21, and 28 post-tumor challenge, mice received i.p. injections of PBS, irrelevant rat IgG (50 μg/mouse), anti-CD40 antibody (FGK4.5; BioExpress; 50 μg/mouse), or poly(inosinic-cytidylic acid) [poly(I:C); Invivogen; 100 μg/mouse] in combination or alone or one dose 21 days post-tumor challenge unless otherwise stated. Flagellin was purified as described previously (22).

For visualization of tumor burden, mice were injected with 0.2 mL of 15 mg/mL luciferin (Promega). After 10 min, animals were anesthetized with isoflurane and imaged using the IVIS 200 system (Xenogen).

We obtained the CP-870,893 monoclonal antibody from Pfizer.

Ascites from mice bearing ID8-Defb29/Vegf-A for 31 days was sorted for CD45⁺CD11c⁺MHC-II⁺ dendritic cells [tumor-derived dendritic cells (tDC)] and used in the assay. Bone marrow–derived dendritic cells (BMDC) or tDC were cultured with 50 μg/mL full-length ovalbumin (Sigma) at 10⁶/mL for 3 h. Negatively bead selected CD3⁺ OT-I splenocytes (1 × 10⁵) were then CFSE labeled and added to ovalbumin-pulsed BMDC or ovalbumin-pulsed tDC (sorted from tumor ascites) in a 10:1 ratio. Either unpulsed tDC or bovine serum albumin (50 μg/mL)–pulsed BMDC were introduced into BMDC/ovalbumin + CFSE OT-I cocultures at various ratios. (At day 5, the total T-cell count was ∼4 × 10⁵). We then collected 30,000 events to detect CFSE dye dilution peaks.

In vivo activated dendritic cells were sorted from either the peritoneal cavity or combined, inguinal, axillary, and brachial lymph nodes of tumor-bearing mice. Quantitative colorimetric arginase determination was done using an Arginase Activity Detection kit (BioAssay Systems). Briefly, 0.05 × 10⁶ to 0.25 × 10⁶ cells were washed and lysed for 10 min in 50 μL of 10 mmol/L Tris-HCl (pH 7.4) containing 0.15 mmol/L pepstatin A, 0.2 mmol/L leupeptin, and 0.4% (v/v) Triton X-100. Lysates were then used to complete the assay according to the manufacturer's instructions.

Total cells were either obtained from peritoneal washes or dissociated spleens of treated or untreated ID8-Defb29/Vegf-A tumor-bearing mice. Peritoneal cells were cocultured for 48 h, in coated and blocked ELISPOT plates, among BMDC in a 10:1 ratio, which were previously pulsed (overnight) with doubly treated (Gamma irradiated and UV-treated) ID8-Defb29/Vegf-A cells (10 dendritic cells:1 tumor cell). Splenocytes were first primed in 24-well plates in a 10:1 ratio with BMDC either pulsed with doubly treated tumor cells or with the peptide mesothelin (GQKMNAQAI; New England Peptide) for 7 days. Cells were then collected and restimulated in 96-well ELISPOT plates with specified antigen or tumor-pulsed BMDC (10:1 ratio) for 24 h. All cultures were maintained in complete RPMI containing 10% fetal bovine serum. Analysis was then continued according to the manufacturer's protocol (eBioscience).

Ovaries of mice were collected and embedded in Tissue-Tek, after which 8 μm sections were made from frozen tissue blocks. Slides were then fixed with acetone and washed with PBS. Sections were then blocked using α-CD32 followed by staining with either biotinylated α-CD8 (53-6.7) or α-CD3 (145-2C11; eBioscience) and completion of immunohistochemical procedure according to the manufacturer's instructions (Vector Labs). Slides were then viewed at ×200 and positive cells were quantified using an image analyzing software (NIS-Element Imaging).

Either peritoneal lavages (10 mL PBS) or culture supernatants were used in ELISA assays for interleukin (IL)-2, IL-12 (p70), IFN-α, and IFN-β (eBioscience and PBL Biomedical Lab, respectively) according to the manufacturers' instructions. Tumor necrosis factor-α was detected using a Mouse-5-Plex panel cytokine assay (Bio-Rad) following the manufacturer's instructions. For in vivo analysis, sorted peritoneal cells (10⁶/mL) from treated or control animals were stimulated for 4 h with phorbol 12-myristate 13-acetate/ionomycin (50 ng/1 μg/mL).

Flow cytometry was done on a FACSCanto (BD Biosciences). Sorting was done on a FACSAria sorter (BD Biosciences).

Anti-mouse antibodies. CD45 (30-F11), CD69 (H1.2F3), and CD11c (HL3; all from BD Biosciences) and MHC-II (NIMR-4; eBioscience).

Anti-human antibodies. CD45 (HI30), DEC-205 (MG38), CD11c (B-ly6), CD3 (UCHT1), CD11b (ICRF44), HLA-DR (L234), and CD14 (M5E2; all from Biolegend).

An allophycocyanin-labeled tetramer consisting of K^b folded with GQKMNAQAI peptide was provided by the NIH Tetramer Core Facility.

Differences between the means of experimental groups were analyzed using the Mann-Whitney test and survival was analyzed with the log-rank test, both using the GraphPad Prism 4.0 software.

Division indices, defined as average number of cell divisions that the responding cells underwent, were calculated using FlowJo software.

CD11c⁺MHC-II⁺ dendritic cells from the ovarian carcinoma microenvironment suppress T-cell function but can be activated to become efficient antigen-presenting cells in vitro. To define the regulatory nature of ovarian cancer–associated dendritic cells, we first confirmed that CFSE-labeled, negatively selected, transgenic OT-I splenic T cells proliferated very poorly in response to tDC pulsed with ovalbumin (9), whereas BMDC induced a robust expansion (Fig. 1A). To show the ability of tDC to suppress naive CD8 T-cell proliferation, we next analyzed the proliferation of OT-I lymphocytes incubated with the same ovalbumin-pulsed BMDC in the presence of two different ratios of CD11c⁺MHC-II⁺ dendritic cells sorted from tumor ascites. Notably, the addition of tumor dendritic cells abrogated the expansion of transgenic T cells (Fig. 1B). This was not caused by a potential interference with access of lymphocytes to pulsed dendritic cells, because ovalbumin-pulsed and washed BMDC, mixed with identical ratios of irrelevant protein-pulsed BMDC, also induced strong proliferation of OT-I lymphocytes (Fig. 1B; Supplementary Fig. S1A; P < 0.05 for the division index).

Confirming previous reports (17, 18, 23), ovarian cancer–associated CD11c⁺MHC-II⁺ dendritic cells express significant levels of PD-L1 (Supplementary Fig. S1B) and secreted immunosuppressive vascular endothelial growth factor (ref. 17; data not shown). More importantly, as recently identified in breast and lung tumor models (24, 25), CD11c⁺DEC-205⁺MHC-II⁺ dendritic cells sorted from ovarian cancer locations, but not dendritic cells sorted from the non-tumor-draining lymph nodes of the same animals, showed strong l-arginase activity (Fig. 1C; P = 0.05). Underscoring the role of the tumor microenvironment in the regulatory differentiation of tumor dendritic cells, incubation of BMDC in tumor conditioned medium increased their production of functional l-arginase (Supplementary Fig. S1C).

Previous reports identified a synergistic effect for combined CD40 and TLR agonists, with combinatorial CD40/TLR3 agonists producing one of the best T-cell responses in healthy mice (15). Because ovarian cancer–associated dendritic cells express detectable levels of CD40 (Supplementary Fig. S1D) and TLR3 (Supplementary Fig. S1E), we next determined whether the synergistic stimulation of CD40 plus TLR3 could reverse the regulatory phenotype of tumor dendritic cells. As shown in Fig. 1D, tumor-derived, ovalbumin-pulsed CD11c⁺MHC-II⁺ dendritic cells induced a significant proliferation of transgenic OT-I lymphocytes only after ex vivo activation with agonistic anti-CD40 antibody plus the TLR3 agonist poly(I:C). Notably, the combined i.p. administration of CD40/TLR3 agonists to ovarian cancer–bearing mice, but not individual treatments, induced a significant decrease in the immunosuppressive l-arginase activity of tumor dendritic cells (Fig. 1C; P = 0.05).

Taken together, these data indicate that ovarian cancer–associated dendritic cells suppress T-cell responses, but their immunostimulatory capacity can be promoted by the synergistic stimulation of CD40 and TLR3.

Synergistic stimulation of CD40 and TLR3 activates mouse and human ovarian cancer–infiltrating dendritic cells. To confirm activation of tumor-associated dendritic cells by concurrent stimulation of CD40 and TLR3, we first treated ascites from ID8-Defb29/Vegf-A tumor-bearing mice with agonistic anti-CD40 plus poly(I:C), the individual treatments, or PBS. As shown in Fig. 2A, combinatorial treatment induced a stronger up-regulation of CD80, CD86, CD70, and MHC-II (P < 0.05) in tumor CD11c⁺ dendritic cells after 24 h compared with incubation with anti-CD40 or poly(I:C) alone. Correspondingly, combined anti-CD40 plus poly(I:C) treatment of CD11c⁺MHC-II⁺ dendritic cells sorted from ovarian cancer locations induced a significant increase in the production of tumor necrosis factor-α (Fig. 2B). Most importantly, sorted peritoneal dendritic cells from ovarian cancer–bearing mice treated with combined CD40/TLR3 agonists secreted significantly higher amounts of immunostimulatory IFN-β and IL-12 (p70) in response to phorbol 12-myristate 13-acetate/ionomycin compared with tumor dendritic cells from mice receiving individual treatments (Fig. 2B; P = 0.05).

To define the applicability of our findings, we stimulated single-cell suspensions from mechanically dissociated stage III to IV human ovarian carcinoma specimens (17) for 48 h with the fully human IgG₂ monoclonal antibody, CP-870,893 (Pfizer; refs. 12, 26). As shown in Fig. 2C, CD80 and MHC-II were up-regulated in some specimens when poly(I:C) was simultaneously added. Therefore, although the synergistic stimulation of CD40 and TLR3 on tumor-associated dendritic cells using currently available human antibodies may be weaker than that of our murine ovarian cancer model, coadministration of CD40 plus TLR3 agonists could also promote the maturation of human ovarian cancer–associated dendritic cells in selected patients.

Concurrent stimulation of CD40 and TLR3 enhances the capacity of tumor-associated CD11c⁺ dendritic cells to process antigens that they spontaneously engulf in the tumor microenvironment. Efficient antigen presentation requires competent antigen engulfment and processing. To confirm the capacity of tumor-infiltrating dendritic cells to spontaneously take-up tumor materials in vivo, we first analyzed the ascites of C57BL/6 mice bearing i.p. ID8-Defb29/Vegf-A engineered to express GFP (9). Up to 18% of total CD11c dendritic cells at this temporal point (day 35) coexhibited green fluorescence, suggesting recent engulfment of tumor materials (Fig. 3A). Then, to define the effect of costimulation of CD40/TLR3 on tumor-associated dendritic cells, on their capacity to take up and process antigens, we treated mice bearing highly aggressive i.p. ID8-Defb29/Vegf-A tumors (9) with three weekly i.p. injections of agonistic αCD40 antibody plus poly(I:C). To monitor antigen processing, 6 days after the last treatment, we administered ovalbumin conjugated to a self-quenched fluorophore (DQ-Ova; Invitrogen), which emits bright green fluorescence only if proteolytic digestion occurs in the endosome (27). As expected, the proportion of DQ-Ova⁺CD11c⁺ dendritic cells that took up and efficiently processed antigen at tumor locations was significantly increased (≈10-fold) in mice receiving combinatorial treatment 24 h before (Fig. 3B; Supplementary Fig. S2A; P < 0.05). Supporting their in situ activation by CD40/TLR3 costimulation, CD11c⁺ dendritic cells at tumor locations in treated mice also expressed higher levels of CD80, CD40, and MHC-I (Fig. 3B; Supplementary Fig. S2B). Correspondingly, the percentages and total number of CD11c⁺ dendritic cells processing GFP⁺ tumor materials and the up-regulation of costimulatory CD80 in ID8-Defb29/Vegf-A-GFP tumor-bearing mice was also significantly higher in treated mice compared with controls (Fig. 3C; Supplementary Fig. S2C). Jointly, these results indicate that coactivation of CD40 and TLR3 dramatically enhances the capacity of tumor-infiltrating dendritic cells to process the antigens that they spontaneously phagocytose at tumor locations and induces the up-regulation of costimulatory determinants.

Costimulation of CD40 and TLR3 in vivo promotes the migration of activated dendritic cells carrying antigens from tumor to lymphatic locations and their capacity to present antigens. Efficient antigen presentation in vivo also requires migration of mature dendritic cells to lymphatic locations. As we have established that synergistic stimulation of CD40 and TLR3 is the most efficacious regimen for dendritic cells activation compared with single treatments, we next sought to define its effect on the migration of activated dendritic cells from tumor locations to draining lymph nodes. We first challenged mice with flank ID8-Defb29/Vegf-A tumors, as described (9), and when tumors reached a size of 200 mm², mice were shaved and painted with 50 μL of 0.1% FITC dissolved in 50:50 (v/v) acetone-dibutylphthalate (28) followed by an intratumoral treatment of CD40/TLR3 agonists or PBS. As expected, the percentage of activated (CD80⁺MHC-II⁺) CD11c⁺ dendritic cells carrying FITC from the tumor location to its draining (inguinal) lymph node increased 2-fold in treated mice compared with controls (Fig. 4A).

There was also a substantial increase in the proportion of CD11c⁺FITC⁺CD80⁺MHC-II⁺ dendritic cells in the auxiliary lymph node, where the inguinal lymph node drains, in treated mice, compared with controls (Fig. 4B).

To confirm the capacity of CD40/TLR3-matured dendritic cells to activate T cells, we sorted CD45⁺CD11c⁺MHC-II⁺ dendritic cells from the peritoneal cavity of treated and control ID8-Defb29/Vegf-A tumor-bearing mice and incubated them with CFSE-labeled allogeneic splenic CD3⁺ negatively selected from BALB/c mice. Dendritic cells sorted from mice receiving PBS did not induce the expansion of allogeneic T cells; in contrast, dendritic cells derived from the ascites of αCD40/poly(I:C)-treated mice elicited significant proliferative responses after 5 days in culture (Fig. 4C). Finally, ELISPOT analysis revealed that the number of negatively immunopurified T-cell splenocytes from tumor-bearing mice, producing granzyme B in response to directly sorted (unpulsed) tumor-associated dendritic cells, was significantly higher when tumor dendritic cells were procured from mice synergistically treated in vivo compared with mice receiving individual treatments (Fig. 4D; P < 0.05). Therefore, synergistic costimulation of CD40 and TLR3 enhances the migration of immunostimulatory dendritic cells carrying local antigens from the immunosuppressive tumor microenvironment to T-cell–rich lymphoid organs and also enhances antigen presentation and T-cell activity.

Administration of CD40/TLR3 agonists to ovarian cancer–bearing hosts boosts T-cell–mediated antitumor immunity. We hypothesized that, if synergistic costimulation of CD40 and TLR3 promotes the capacity of tolerogenic dendritic cells from tumor locations to efficiently present the tumor antigens that they carry to lymph nodes, enhanced tumor-specific T-cell responses should become measurable. Supporting this proposition, mice bearing established ID8-Defb29/Vegf-A tumors treated with combinatorial CD40/TLR3 agonists contained 3-fold more antigen-experienced (CD44⁺) CD8⁺ T cells in peritoneal wash samples compared with mice receiving irrelevant IgG or PBS (Fig. 5A). Most importantly, the number of peritoneal T cells producing IFN-γ in response to tumor antigens in ELISPOT analyses was significantly increased in ID8-Defb29/Vegf-A ovarian cancer–bearing mice treated with combinatorial anti-CD40 agonistic antibody plus poly(I:C) compared with control mice receiving PBS (Fig. 5B; P < 0.05). Further confirming the synergistic effect, individual treatment with either anti-CD40 antibodies or poly(I:C) did not result in any measurable increase in the number of T cells producing IFN-γ on stimulation with tumor antigens (Fig. 5B). Comparable results were found using splenic T cells (Fig. 5B; P < 0.05). Finally, the percentage of tumor antigen-specific splenic CD8⁺ T cells exhibiting a central memory (CD44⁺CD62L⁺) phenotype, specifically recognizing a H-2Db–restricted mesothelin epitope expressed by ID8 tumor cells (29) in tetramer analyses, was significantly higher in mice treated with combinatorial CD40/TLR3 agonists compared with mice receiving separate agonists or PBS (Fig. 5C; Supplementary Fig. S2D; P < 0.05; although we did not detect differences in tetramer⁺CD8⁺ among groups). This result was corroborated by the enhanced secretion of IFN-γ from combinatorial treated splenocytes in response to the tumor peptide, mesothelin (Fig. 5C; P < 0.05).

Consistent with a dramatic enhancement of tumor-reactive T-cell–mediated responses elicited by in situ administration of CD40 plus TLR3 agonists, the infiltration of tumor islets in resected ID8-Defb29/Vegf-A tumor-bearing ovaries by CD8⁺CD3⁺ cytotoxic T cells (the only known cell type in the ovarian cancer microenvironment that elicits immune pressure against tumor progression; ref. 7) was significantly stronger in samples from mice treated with combinatorial CD40/TLR3 agonists (Fig. 5D; Supplementary Fig. S2E; P < 0.05). Collectively, these data indicate that the immunostimulatory phenotype promoted by CD40/TLR agonists on otherwise immunosuppressive tumor-infiltrating dendritic cells is associated with the dramatic expansion and activation of tumor-reactive T cells.

Synergistic stimulation of CD40 and TLR3 induces the rejection of otherwise lethal i.p. ovarian carcinomas. As we have shown that the synergistic use of CD40 and TLR3 agonists elicits enhanced antitumor immunogenic boosts, we next investigated its therapeutic potential. We first challenged mice with i.p. ovarian carcinomas, developed with parental ID8 cells transduced with luciferase, for intravital monitoring of tumor burden (17). Mice growing established tumors received weekly combinatorial CD40/TLR3 agonists, rat IgG, or PBS for 4 weeks, with no evidence of toxicity in any group. Notably, in the absence of direct targeting of tumor cells, administration of CD40/TLR3 agonists resulted in the elimination of any obvious disease in ≈50% of mice, which remained healthy >300 days after tumor injection, whereas all mice receiving control injections succumbed to the disease ≈190 days earlier (Fig. 6A). Correspondingly, synergistic treatment significantly reduced tumor burden as determined by luminescence 68 days post-tumor injection (Fig. 6B; Supplementary Fig. S2F).

To confirm the therapeutic potential of combinatorial treatment against an even more aggressive ovarian carcinoma model, we also treated mice growing established i.p. ID8-Defb29/Vegf-A tumors, which accelerates tumor progression by ∼3-fold (9). In the absence of complementary therapies, synergistic stimulation of CD40 and TLR3 also induced a significant 15% increase in lifespan with only four injections (Fig. 6C; P < 0.001). Notably, 90% of mice treated with CD40/TLR3 agonists were alive at the time (51 days) when all control mice succumbed to tumor malignancy. Additionally, when we administered a TLR5 agonist with αCD40 for treatments, we saw that its effect was inferior to the αCD40/poly(I:C) combination (Supplementary Fig. S2G; P < 0.01), although ovarian cancer–associated dendritic cells express TLR5 (23). Interestingly, CD40/TLR3 agonists treatments induced an astonishing 450-fold increase in IFN-α levels in peritoneal wash supernatants, accompanied by a ∼2.5-fold increase in IL-12 (p70) and a significant increase of IL-2 (Fig. 6D; P < 0.05). Therefore, synergistically activating CD40 plus TLR3 in hosts bearing established ovarian cancer results in a significant therapeutic benefit even in the absence of additional surgical or chemotherapeutic interventions and, depending on the tumor, induces durable curative results.

Ovarian cancer is one of the most aggressive and frequent forms of epithelial cancer and claims >15,000 lives every year in the United States alone (3). The most abundant leukocyte subset in the microenvironment of human and mouse solid ovarian carcinoma specimens is CD11c⁺DEC-205⁺MHC-II⁺CD11b⁻ dendritic cells with proangiogenic and immunosuppressive activity (9, 10, 17). Here, we show that the synergistic stimulation of CD40 and TLR3 transforms tumor-associated mouse dendritic cells in vivo from an immunosuppressive to an immunostimulatory cell type that efficiently processes spontaneously engulfed tumor antigens up-regulates costimulatory molecules and migrates to lymphatic locations to activate antigen-specific T cells. Consequently, the administration of CD40/TLR3 agonists to established ovarian cancer–bearing hosts boosts T-cell–mediated antitumor immunity, resulting in the rejection of otherwise lethal i.p. ovarian carcinomas.

Despite inducing measurable immune responses in most cancer patients, vaccination approaches using ex vivo conditioned dendritic cells have thus far induced limited therapeutic effects (1, 2). The success of dendritic cell–based immunotherapy in stimulating antitumor cellular immunity is decidedly dependent on trafficking of mature dendritic cells to T-cell–rich lymphoid organs after tumor antigen processing. Ovarian cancer–infiltrating dendritic cells spontaneously engulf tumor cells, avoiding the need to prime them ex vivo or in vivo with exogenous tumor antigens. CD40/TLR triggering enhances both antigen processing and lymphatic migration, which results in increased numbers of tumor-specific T cells with central memory (CD44⁺CD62L⁺) attributes in the spleen, in conjunction with IFN-γ–secreting T effector cells in spleens and at tumor locations. Although the phenotypic transformation induced by combined CD40/TLR3 agonists on human ovarian cancer–infiltrating dendritic cells was clearly weaker than that of our mouse model, we also observed that CD80 and MHC-II were up-regulated in some specimens. Therefore, combining CD40 and TLR3 agonists, by promoting the immunostimulatory potential of immunosuppressive/proangiogenic tumor antigen-presenting dendritic cells in situ in selected patients, could enhance the therapeutic effect of these reagents, currently being individually tested (12, 30, 31).

We have shown previously that the elimination of immunosuppressive/proangiogenic dendritic cells from ovarian cancer locations results in measurable therapeutic effects (17). Rather than depleting them, this new approach transforms tumor-infiltrating dendritic cells into ”Trojan Horses,“ resulting in stronger therapeutic benefits. As preliminary reports indicate that the persistence of antitumor T cells adoptively transferred into ovarian cancer patients may be significantly reduced by tumor microenvironmental factors, it will be therefore interesting to determine whether eliminating this abundant immunosuppressive component from the ovarian cancer microenvironment while boosting endogenous antitumor immunity enhances the persistence and/or the effectiveness of T-cell adoptive therapies in a future clinical setting.

C.L. Ahonen: cofounder, stockholder, and consultant, ImmuRx. The other authors disclosed no potential conflicts of interest.

Grant support: 2006-2011 Liz-Tilberis Award and National Cancer Institute grant RO1CA124515; NIH training grant T32AI007363 (U.K. Scarlett).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank the NIH Tetramer Core Facility for providing the tetramer, R.J. Noelle for providing reagents and critical expertise, and Pfizer for providing the human anti-CD40 Ab.