Topical application of tumors with the TLR7 agonist imiquimod is an effective adjunct treatment for a range of primary dermatological cancers. However, for therapy to be effective against a broad range of solid tumor types, it must promote a strong systemic antitumor response that targets metastases in addition to primary tumor. We therefore investigated the potential of locally delivered imiquimod to stimulate an effective systemic antitumor response in a murine model of malignant mesothelioma (AB1-HA) with primary and distal tumors (dual tumor). Persistent delivery of imiquimod into primary tumor significantly retarded tumor growth in all treated mice compared with vehicle control. This local antitumor immune response required both CD8 T cells and NK cells, but not CD4 T cells, and was reliant on type I IFN induction. In vivo CTL studies and Ly6A/E staining of lymphocytes suggested that local imiquimod treatment had indeed induced a systemic, Ag-specific CD8 response. However, notably this response was not sufficient to retard the growth of an untreated distal tumor. Because local imiquimod treatment did not induce significant CD4 T cell responses, we investigated the efficacy of combining imiquimod with agonistic CD40 Ab (as a surrogate for CD4 T cell help). Combination of locally delivered imiquimod with systemic anti-CD40 immunotherapy not only significantly enhanced the local antitumor response, with 30% complete resolution, but it was also effective at significantly retarding growth of distal tumor. These results demonstrate that antitumor responses induced by locally delivered TLR7 agonists can be harnessed systemically for treating distal tumor.
The immune system can recognize tumor Ags and elicit CD8 T cell responses against tumor Ags through the process of cross-presentation (1). However, cross-presentation of tumor Ags is often associated with weak or insufficient priming of CD8 T cell responses that are incapable of resolving tumors or even slowing their progression (2, 3). This may relate to the many identified suppressive mechanisms (such as T regulatory cells, TR1 cells, PD1, etc.) that are able to suppress effector CD8 T cell function. Alternatively, it may be due to events during the initial encounter with tumor Ag, leading to an absence of a signal 3 required for full arming of CD8 responses (4). In both cases, it has become clear that engaging the innate immune system, through activation of pattern recognition receptors such as the TLR family, is required to provide the appropriate context to allow effective CD8 T cell responses to be generated and function.
TLRs are expressed in a wide range of different cell types (5). Most are highly expressed in key immune cells such as dendritic cells (DCs),3 macrophages, and T cells, but expression has also been detected in some tumors (6). TLR stimulation of DCs is known to switch their activation state from an immature, nonimmunogenic, and tolerogenic phenotype to a mature condition, capable of cross-presenting Ag to elicit effective CD8 T cell responses (7, 8, 9). DCs also respond to cytokines and other signals from surrounding immune cells as indirect stimuli (10) and are known to work in concert with NK cells following TLR ligation to stimulate bystander T cell activation, via IFN-αβ production and stimulation of NK cells to secrete IFN-γ (11). We have recently screened various TLR ligands for antitumor efficacy and identified a TLR7 ligand (3M-019, an analog of imiquimod) to be effective in delaying tumor growth (12). Because imiquimod itself has been commercially tested and is readily available for translation to clinical use, we decided to further scrutinize its efficacy in vivo in a representative murine model of treatment-resistant solid malignancy.
Imiquimod is a small-molecule immunomodulatory compound of the imidazoquinoline family that exhibits both antiviral and antitumor properties. Its antitumor properties are known to be at least 3-fold: proapoptotic, antiangiogenic, and immune stimulating (13, 14, 15). As a nucleoside analog that mimics the immune response to viral ssRNA, imiquimod exerts its biological efficacy through agonistic stimulation of TLR7 in immune cells such as DCs (16, 17), resulting in the production of proinflammatory cytokines and chemokines. In particular, the substantial IFN-α production by plasmacytoid DCs induces monocyte differentiation into TLR7 expressing DCs, thereby amplifying IFN-α production (18). Commercially, imiquimod has been effectively used as a topical agent (Aldara) for therapeutic treatment of a number of dermatological tumors, including basal cell carcinomas (19, 20), intraepidermal keratinocyte neoplasias (21), and cutaneous metastases of melanoma (22, 23). Imiquimod has also been investigated in some animal models of tumors (24, 25), with varying degrees of efficacy. In the majority of these scenarios, imiquimod has been used to target localized tumors, with the aim of destroying well-delineated and accessible primary tumors, with a focus on the cytotoxic capabilities of the drug, rather than the immune priming activity. However, based on its powerful innate activating properties, imiquimod should also be able to initiate and sustain systemic immune responses, including tumor-specific CD8 T cell responses.
In this study, we tested the hypothesis that persistent treatment with a TLR7 ligand provides an immunostimulatory context to cross-presented tumor Ags capable of inducing systemic CD8 T cell responses. Using a mouse model of mesothelioma that is transfected with influenza virus hemagglutinin (HA) as a cross-presented neoantigen (26), we identify some of the key cellular factors required for effective local imiquimod responses (namely CD8 T cells, NK cells, and type I and II IFN) and some of the potential factors limiting responses in distal tumors (specifically CD4 T cell activation and CD40-driven costimulation).
Materials and Methods
High purity imiquimod (98%) was purchased from Calbiochem. CFSE was purchased from Molecular Probes. HA peptide (IYSTVASSL) was synthesized by the Centre for Cell and Molecular Biology (University of Western Australia, Perth, Australia).
Agonistic mAb to mouse CD40 receptor (FGK45) was purchased from the Monoclonal Antibody Facility (K. Davern, Western Australian Institute for Medical Research). Mice received one dose of 100 μg of Ab i.v. from day 1 of imiquimod treatment for a total of three doses over the course of treatment (q2d × 3). Murine specific mAbs either unconjugated or coupled to FITC, along with appropriate matching isotype controls, were purchased from BD Pharmingen and eBioscience as follows: CD4 (RM4-5), CD8 (53-6.7), Ly6A/E (D7), TCR-β (Η57-597), and CD49b (DX5). Flow cytometry was conducted using a BD Biosciences FACSCalibur instrument and analyzed using CellQuest software (BD Immunocytometry Systems).
BALB/c (H-2d) wild-type and nude (BALB/c background) mice were purchased from the Animal Resources Centre (Canning Vale, Western Australia) and maintained under specific pathogen-free conditions. HA-specific TCR transgenic mice (CL4 mice, BALB/c background) that express a TCR specific for the H-2d-restricted peptide IYSTVASSL (residues 518–526) of A/PR8/8/34 (H1N1) influenza virus HA were generated and screened, as described previously (27). BALB/c mice doubly deficient for perforin and IFN-γ (BALB/c.pfp−/−.IFN-γ−/−) were produced and bred at the Peter MacCallum Cancer Centre (28). All experiments used female mice between 6 and 8 wk of age. Animal experimentation was conducted according to University of Western Australia Animal Ethics Committee approval 03/100/434.
Tumor cell culture and inoculation
The mouse (BALB/c)-derived mesothelioma cell line AB1-HA was previously generated by transfection of the parental AB1 line with the gene encoding influenza HA protein (AB1-HA), as described by Marzo et al. (27). Cell cultures were maintained in RPMI 1640 (Life Technologies), supplemented with 20 mM HEPES, 0.05 mM 2-ME, 60 μg/ml penicillin (CSL), 50 μg/ml gentamicin (West), and 5% FCS (Life Technologies). Geneticin (Life Technologies) was added to medium at a final concentration of 400 μg/ml to select for AB1-HA transfectants. HA expression on transfected cells was confirmed by FACS analysis using biotinylated HA-specific mAb H18 (29), originally obtained from W. Gerhard (Wistar Institute, Philadelphia, PA). Tumor cells were injected s.c. into the shaven right flank (or both flanks for dual tumor experiment) of female BALB/c mice on day 0 (1 × 106 cells in 100 μl of injectable saline), and tumor growth was monitored by taking two perpendicular diameter measurements using microcallipers. Mice were euthanized when tumors reached 100 mm2.
Intratumoral treatment for BALB/c wild-type mice commenced on day 10 when tumors were palpable (∼4 mm2). Treatment of tumor-bearing BALB/c nude mice commenced on day 7 (equivalent in size to day 10 in wild-type mice). Mice were lightly restrained, and the surface of the tumor was swabbed with alcohol before direct injection of the ligand diluted in 50 μl of sterile injectable saline (Baxter Healthcare). Care was taken to slowly deliver the bolus, and the base of the tumor was clasped firmly to prevent liquid penetrating through the tumor. Tumors were observed to swell slightly and blanch upon injection. Tumors were injected in a different site for each treatment day. Various multiple dosing regimens and concentrations were trialed.
Immune cell depletion studies
NK cell depletion was performed using anti-asialo-GM1 Ab (Wako Fine Chemicals). A total of 20 μl of anti-asialo-GM1 Ab diluted with 180 μl of sterile saline was injected i.p. 1 day before imiquimod treatment and every 3 days thereafter for a total of four doses. NK cell depletion (>90%) was verified during treatment by FACS analysis of peripheral blood using Abs specific for TCRβ and CD49. CD4, CD8α, and CD4+ CD25+ T cell depletion was performed using the purified GSK1.1, YTS.169, and PC61 mAbs, respectively (prepared by K. Davern, mAb Facility, Western Australian Institute for Medical Research). Mice received an initial dose of 200 μg i.v., 1 day before imiquimod treatment, followed by a second dose of 150 μg administered i.p. on the day of treatment and then 150 μg i.p. every second day thereafter for a total of six doses. CD8 depletion (>95%) was verified during treatment by FACS analysis of peripheral blood using Abs specific for TCRβ and CD8α.
Mouse IFN-α ELISA was performed according to the manufacturer’s protocol (PBL Biomedical Laboratories). The plates were read at 450 nm by Victor2 V 1420 Multilabel Counter (PerkinElmer Life Sciences). Serum from three separate animals in each group was tested and assayed in triplicate.
In vivo IFN-αβ blockade
The origin, purification, and assay of IFN-αβ-neutralizing sheep Ig (anti-IFN-αβ) and matching normal sheep Ig have been previously described in detail (30, 31). Sheep IgG or anti-IFN-αβ was delivered, as described (32). In brief, mice were injected i.v. with 0.2 ml of Igs on day −1, +2, and + 4 with respect to imiquimod administration. Neutralization was confirmed by staining for Ly6A/E expression on peripheral blood CD4 and CD8 T cells 24 h after the first dose of imiquimod.
Preparation and staining of tissues for flow cytometry
For flow cytometry analysis, tumors, spleens, and lymph nodes were removed from mice and placed into ice-cold PBS containing 1% (v/v) FCS. The axillary and inguinal nodes were pooled for the tumor flank (draining lymph nodes) and for the contralateral flank (nondraining lymph nodes) in all cases. Tissues were homogenized into cell suspensions using frosted glass slides and washed twice with cold PBS with 1% (v/v) FCS and 0.01% (w/v) NaN3 (wash buffer) before being resuspended in FACS buffer (PBS with 1% (w/v) BSA and 5% (v/v) FCS). After 30 min, cells were washed thrice in wash buffer and resuspended in 2% (v/v) formalin/PBS (Sigma-Aldrich) before FACS analysis.
In vivo CTL assay
Investigation of HA-specific in vivo CTL killing of target cells was performed, as described previously (33). Briefly, BALB/c spleens were depleted of erythrocytes by resuspending the cells in RBC lysis buffer for 5 min, followed by three washes in PBS and resuspension at a concentration of 5 × 106 cells/ml. Cells were split into two equal populations as follows: one was pulsed with 1 μg/ml HA peptide (IYSTVASSL) for 90 min at 37°C; the other was treated the same way, but without addition of peptide. Cells were subsequently labeled with CFSE for 10 min at room temperature at a final concentration of 5 μM CFSE (peptide-pulsed cells; CFSE high) and 0.5 μM CFSE (unpulsed cells; CFSE low). CFSE high and low cells were washed with FCS four times, followed by PBS, before being pooled in equal proportions and injected i.v. into recipient mice (1 × 107 cells in 300 μl of saline per mouse). After 18 h, the draining and nondraining lymph nodes and spleens were harvested as above and analyzed by FACS for recovery of CFSE high and low cells. The percentage of killing was determined as follows: 100 × (1 − (CFSE high events/CFSE low events)).
Surface Ags were detected using the streptavidin-biotin-labeling immunoperoxidase-staining technique. Tumors were removed, placed in compound-embedding medium (OCT; Miles), snap frozen using dry ice, and stored at −80°C. Ten-micrometer sections were cut, collected on poly(l-lysine)-coated slides, and allowed to air dry. Slides were stored at −20°C over desiccant before staining. Before immunostaining, sections were fixed with cold ethanol (15 min) and blocked with 1% (v/v) H2O2 (5 min), followed by avidin/biotin block (Vector Laboratories) (10 min each). Sections were incubated with the appropriate dilutions of primary rat anti-mouse mAbs against CD8 and CD4, or isotype controls for 1 h, followed by incubation with a biotinylated secondary Ab for 30 min (mouse anti-rat IgG F(ab′)2; Jackson ImmunoResearch Laboratories). Immunostaining was detected by incubating with streptavidin-HRP (DakoCytomation) for 30 min and with diaminobenzidine-H2O2 (Sigma-Aldrich) for 5–10 min. Slides were washed three times for 5 min each time in PBS between each incubation step, counterstained with hematoxylin, and mounted in aqueous mounting medium.
Data were statistically evaluated using Prism software (GraphPad). Survival responses were analyzed by Kaplan-Meier using the log-rank test. Growth curves were compared using two-tailed paired t test, with pairs defined by time point. All other variables were compared using a two-tailed Mann-Whitney U test. Significance was defined as p < 0.05.
Locally delivered imiquimod significantly retards the growth of AB1-HA tumors
We initially screened the antitumor activity of imiquimod using a number of dosing regimens (q2d × 3, q2d × 6, q1d × 3) and doses (1 μg to 1 mg) delivered through various routes (i.p., s.c., i.v., topical, peritumoral, and intratumoral) (data not shown). Only locally (intratumoral) delivered imiquimod consistently and significantly retarded the growth of AB1 and AB1-HA in comparison with vehicle (citrate-buffered saline (CBS)) (Fig. 1). Optimal responses were observed when treatment was given daily (q1d × 6) at a dose of 50 μg/ml. Maximum inhibition of tumor growth resulted in a >50% increase in survival time (p < 0.05) and was not associated with overt signs of toxicity (such as weight loss and ruffled fur). This dosing regimen was also effective against other tumor types, including renal cell carcinoma and 4T1 breast cancer, with similar benefits in survival time (data not shown).
Systemic type I IFN is an important component in the response to imiquimod
TLR7 ligands are strong inducers of type I IFN, and their antiviral and antitumor properties have been largely associated with this phenomenon (24, 34, 35). To determine whether systemic type I IFN could be a factor driving the immune response to imiquimod in our AB1-HA model, we first examined whether local imiquimod could indeed induce systemic release of IFN. To do this, we measured the expression of Ly6A/E on peripheral blood CD4 and CD8 T cells because we and others have found this to be a sensitive marker of type I IFN exposure even when levels are below detection by ELISA (32). Staining of whole blood after three doses of imiquimod revealed a positive shift in expression of Ly6A/E on CD8+ lymphocytes, but not on CD4+ lymphocytes (Fig. 2,A). To confirm this systemic response, we measured the presence of IFN-α in the sera of three tumor-bearing mice treated for 3 days with either imiquimod or vehicle by ELISA. Imiquimod treatment of tumors stimulated the systemic secretion of IFN-α as detected in the serum (5.05 pg/ml); this was significantly greater than for vehicle-treated mice (0.72 pg/ml) and naive (tumor-free) mice (0.46 pg/ml; p < 0.05). To evaluate the importance of this type I IFN response in the antitumor effects of imiquimod, we treated tumor-bearing mice with IFN-αβ-blocking serum before and throughout treatment with imiquimod. Control mice were treated with vehicle, nonspecific serum, and imiquimod, or IFN-αβ-blocking serum alone. Neutralization of IFN-αβ did indeed inhibit the antitumor effect of imiquimod, with no signficant antitumor effect observed in the presence of the blocking serum (Fig. 2 B).
The antitumor immune response to imiquimod requires CD8 T cells
Imiquimod is known to act as an immune response modifier and as a proapoptotic agent, and it has antiangiogenic properties. To identify the extent to which the immune system was involved in the antitumor response and the cellular components involved, nude mice lacking functional T cells were inoculated with the AB1-HA tumor and treated using the identical protocol to BALB/c mice. The lack of T cells totally abrogated the antitumor effect of imiquimod (Fig. 3,A), with survival no different from that of vehicle-treated mice. To further investigate the importance of select immune subsets in the antitumor response to imiquimod, BALB/c mice bearing AB1-HA tumors were depleted of CD8, CD4, NK, and CD4+CD25+ regulatory T cells before and throughout treatment with imiquimod. Depletion of either CD8 T cells or NK cells completely abrogated the efficacy of imiquimod, with survival time in these mice equivalent to undepleted mice treated with vehicle (Fig. 3 B). In contrast, CD4 depletion marginally enhanced the activity of imiquimod with a 20% increase in mean survival time in comparison with undepleted mice treated with imquimod, suggesting the potential for CD4-mediated suppression in this system. However, depletion of CD25+ cells using PC61 Ab, which has been shown to deplete CD4+CD25+ regulatory T cells, had no impact on the response to imiquimod.
TLR7 ligation stimulates a systemic CD8 effector T cell response
Having demonstrated that the antitumor response induced by imiquimod required both type I IFN and CD8 T cells, we next measured the systemic killing ability of tumor-specific CD8 T cells using the in vivo CTL assay (Fig. 4,A). Type I IFNs have been shown to promote the expansion and function of CD8 T cells (20) and immune cell responses to cross-presented Ags (36), and also stimulate CD8 T cell expansion and function (37). In keeping with this, we found that imiquimod treatment promoted a >3-fold increase in the killing of HA-specific targets in tumor-draining lymph nodes (up to 100% in some individuals) in comparison with vehicle treatment (Fig. 4 B). Moreover, this enhanced killing response was sustained systemically, being observed in the spleen and nondraining lymph nodes. The local tumor site was also investigated; however, insufficient cell numbers were recovered for meaningful and accurate analysis.
The killing of peptide-pulsed target cells is often used as a surrogate for measuring in vivo effector responses. However, tumor cells are the true target of such responses and can display considerable resistance to killing (20, 27). Thus, to determine whether the local antitumor response induced by imiquimod indeed required enhanced effector responses, we tested the effectiveness of imiquimod in BALB/c perforin/IFN-γ double gene-targeted mice. We elected to use double gene-targeted mice for these experiments because perforin/granzyme and IFN-γ have been implicated both in combination and alone in mediating effective CD8 effector responses. The effect of imiquimod was completely abrogated in BALB/c-pfp−/−/IFN-γ−/− mice with responses equivalent to vehicle-treated wild-type BALB/c mice (Fig. 5 A), and not significantly different from tumor growth in untreated pfp−/−/IFN-γ−/− mice.
Thus, imiquimod induces potent systemic tumor-specific effector responses, and local inhibition of tumor growth requires effector activity. However, the fact that the antitumor activity of imiquimod was relatively modest suggested that there may be a limit on either the quantity or quality of the effector response at the tumor site. We therefore examined the extent to which locally treated tumors were infiltrated by CD4 and CD8 T cells. A dense infiltrate of CD8 T cells was seen in imiquimod-treated tumors (Fig. 5 B), whereas virtually no CD8s were observed in vehicle-treated tumors, suggesting that access to the tumor site by effector cells was not a limit per se on the effectiveness of imiquimod. Notably, there was no observable CD4 T cell infiltrate in either vehicle- or imiquimod-treated tumors.
Anti-CD40 immunotherapy enhances the antitumor immune response to imiquimod
A lack of CD4 T cell infiltrate, combined with the absence of systemic activation of CD4 T cells by imiquimod-induced type I IFN (Fig. 2,A), led us to hypothesize that the effectiveness of antitumor CD8 T cell responses might be qualitatively enhanced by providing or bypassing CD4 T cell help. To test this, we treated mice with imiquimod in combination with a systemically delivered agonistic anti-CD40 Ab (FGK45), because CD40L (CD154) is expressed by activated CD4 T cells, and CD40L-CD40 interactions are important in mediating CD4 Th cell functions for CD8 T cells (38). As before, a significant antitumor effect was observed when mice were treated with imiquimod (median survival time of 33 days vs 18 days in vehicle-treated animals) (Fig. 6 A). Importantly, combined therapy with imiquimod and anti-CD40 was most effective, with 30% of mice having complete resolution and the remaining mice having a significant increase in median survival time in comparison with untreated mice (p < 0.05) and with mice receiving either imiquimod or anti-CD40 alone (41.5 days vs 33 days and 25 days, respectively; p < 0.05).
When we examined the degree of CD8 infiltration in imiquimod-, anti-CD40-, or combined imiquimod/anti-CD40-treated tumors, we found that imiquimod again induced significant infiltrate, whereas anti-CD40 alone had limited effect (Fig. 6 B). Notably, the combined therapy induced an infiltrate equivalent to imiquimod treatment alone, suggesting a qualitative rather than quantitative improvement in antitumor effector responses, or an increase in the proportion of specific antitumor CD8 T cells within the tumor.
Combined imiquimod and anti-CD40 therapy promotes systemic antitumor responses
To determine whether the improvement in antitumor effector responses seen with the combination therapy translated into truly systemic antitumor immunity, we set up a dual tumor model in which the local (right flank) tumor was injected with imiquimod (with or without systemic anti-CD40 therapy) and the distal (left flank) tumor was left untreated. Imiquimod treatment alone showed a nonsignificant improvement in survival time over vehicle-treated mice; however, this did not reflect a resolution of distal tumor or tumor-free survival. As previously observed, the combination treatment significantly delayed progression of tumor locally, and in most cases completely resolved the tumor by day 30 (10 of 11 mice) compared with the distal tumor (6 of 11 mice tumor free) (Fig. 7). These six mice remained tumor free indefinitely and resisted rechallenge with tumor (data not shown). Overall, survival of the combination-treated mice (55%) was significantly improved compared with imiquimod (0%) or anti-CD40 alone (15%) (p < 0.05). Depletion of CD8 T cells completely abrogated the antitumor effect of the combination treatment (Fig. 7 A), with the response equivalent to vehicle-treated animals (p < 0.05). CD4 T cell depletion also significantly reduced survival (p < 0.05), but not to the same extent, with the overall survival response equivalent to imiquimod-treated animals, but less than for anti-CD40-treated animals.
Few immunotherapies can eradicate established solid tumors. One potential strategy is to use a primary site as a vaccination site to induce systemic antitumor responses (39). In this study, we demonstrate that direct TLR7 stimulation in the tumor, using the small-molecule ligand imiquimod, significantly delayed growth of AB1-HA tumors. This antitumor effect of TLR7 ligation was associated with a strong and systemic antitumor CD8 T cell response and a systemic type I IFN response. We found that both were required for control of tumor growth. Also, persistent TLR stimulation was required for optimal antitumor effects, similar to our previous work with the TLR3 ligand poly(I:C) (12), and consistent with previously reported findings (40). However, TLR7 ligation as a therapy was only effective when imiquimod was administered locally. Distal, untreated tumors did not respond significantly to the treatment. Thus, our data imply that a strong and systemic antitumor CD8 T cell response does not a priori translate into tumor regression. The gap between CD8 T cell responses and tumor regression was bridged by CD40 stimulation: when local imiquimod was combined with systemic anti-CD40 Abs, both treated and distal tumors responded to treatment.
The first question raised by our study is how imiquimod generates a systemic antitumor CD8 T cell response. Antitumor T cell responses were evaluated using the in vivo CTL assay, which measures localized in vivo killing of peptide-pulsed target cells. For this, we used the Kd-restricted HA518–526 epitope (IYSTVASSL), which is expressed by the tumor. The measurement of in vivo killing implies that imiquimod drives the differentiation of antitumor CD8 T cells into CTLs, which migrate to the tumor, the spleen, and different lymph nodes. It is likely that type I IFNs play a key role in this process. AB1-HA tumor Ags are cross-presented to CD8 T cells, where they are weakly priming (41). There is strong evidence that type I IFNs are essential for the generation of CD8 T cell responses to cross-presented Ags (36, 42). Type I IFNs have direct effects on CD8 T cells, regulating their expansion and differentiation into memory cells (37). Indeed, in vivo IFN-αβ neutralization abrogated the (CD8 T cell-dependent) antitumor response. Therefore, our data are consistent with a model in which DCs produce IFN-αβ after TLR7 stimulation, which, in turn, stimulates CD8 T cell responses to cross-presented Ags. TLR7 is mostly expressed by plasmacytoid DCs, whereas the CD8α+ DCs that are responsible for tumor-Ag cross-presentation express much less TLR7 (43). It will be interesting to determine whether TLR7 ligation has to occur in the same DCs that cross-present Ag or whether IFN functions in trans (e.g., pDCs produce type I IFNs that stimulate CD8α+ DC-driven CD8 T cell responses). Antitumor CTLs differentiated into memory cells, because cured mice resisted tumor rechallenge, indicating the presence of functional memory. Notably, AB1-HA tumor-bearing mice that were cured through imiquimod treatment also rejected rechallenge with the parental AB1 tumor cell line (that does not express HA), demonstrating that the repertoire of the antitumor CD8 T cell response extended beyond the HA tumor neo-Ag.
The second key question is why a systemic antitumor CTL response is not sufficient to affect the growth of distal tumors. Several lines of evidence indicate that the nature of the antitumor CD4 T cell response could provide an explanation. First, we found that only CD8 T cells up-regulated Ly6A/E in response to IFN-αβ and CD4 T cells did not respond. Second, imiquimod was associated with a strong influx of CD8 T cells, but not CD4 T cells, into the treated tumor. Third, CD40 stimulation, which has been shown to mimic CD4 help, had a clear capacity to convert a systemic CTL response into a systemic antitumor response. Furthermore, anti-CD40 Abs converted the partial antitumor response induced by imiquimod into a curative response. Thus, it is possible that the imiquimod-induced antitumor T cell responses are defective in engaging the CD4 compartment. Interestingly, CD40 stimulation did not result in a further influx of CD8 T cells into the imiquimod-treated tumor. Therefore, we conclude that anti-CD40 Abs either increase the quality of the CD8 T cell response, or the sensitivity of tumor cells to T cell-mediated killing, but not the quantity of the response. Indeed, it has been shown that CD4 T cells are required for the maintenance of CD8 T cell responses, preventing exhaustion (44) and providing a postlicensing signal in the tumor (45). It is therefore possible that anti-CD40 also provides a local postlicensing signal to CD8 T cells directly in the tumor itself. This may occur through effects directly on DCs because the costimulatory and Ag-presenting capacity of APCs is greatly enhanced by ligation of CD40 on the cell surface, and blockade of CD40L negates CTL priming (38). Intriguingly, whereas CD4 depletion had no significant impact on the efficacy of imiquimod in single or dual tumors (data not shown), it did significantly limit the efficacy of the imquimod and anti-CD40 combination, with the overall survival reduced to the level of imiquimod only. This suggests that anti-CD40 might actually be indirectly or directly engaging CD4 help rather than bypassing it. This could involve the release of CD4-activating cytokines from DCs after CD40 ligation, as has been demonstrated for NK cell activation in antitumor responses induced by FGK therapy (46). A direct effect of anti-CD40 on CD4 function cannot be ruled out either, because the presence of functional CD40 on activated CD4 T cells (such as in the autoimmune setting) has been demonstrated (47, 48), with direct ligation leading to costimulatory activity akin to CD28, and release of IL-2. Such a mechanism could act independently of APC activation, adding to the potential APC-activating function of imiquimod.
Alternatively, it is possible that anti-CD40 may exert its effects through mechanisms other than bypassing the requirement for CD4 T cell help. For example, anti-CD40 Abs have strong effects on (tumor) blood vessels. This could alter the tumor stroma, allowing CTLs to kill tumor cells. Ongoing studies are aimed to distinguish between these two possibilities. Nevertheless, our data support the following model: the imiquimod-treated tumor serves as the source of tumor Ag and drives the antitumor CTL response. Differentiation of CD8 T cells into antitumor effectors most likely occurs in the tumor-draining lymph nodes, after which CTLs leave the lymph nodes and migrate toward the tumor, but also to other lymphoid tissues. However, these CTLs can only kill tumor cells in the treated tumor, possibly as a result of local high IFN levels or local MHC class I up-regulation (12). Systemic CD40 stimulation allows CTL to also kill tumor cells in distal tumors that were not exposed to imiquimod, possibly via altering the function of tumor-infiltrating APCs or blood vessels, or other mechanisms. In this model, imiquimod serves to turn the tumor into its own vaccine. These results have important implications for translation. If a single readily accessible tumor mass can be treated with a TLR ligand such as imiquimod, then distal metastases could be sensitized to the ensuing CTL response using anti-CD40 Abs. Importantly, imiquimod is already in clinical use in the form of a topical cream (Aldara), and clinical grade anti-CD40 Abs are currently being trialed. Finally, our data also suggest that the single measurement of antitumor CTL responses as a readout for antitumor vaccines or immunotherapies may have a poor predictive value for therapeutic efficacy.
The authors have no financial conflict of interest.
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.
This work was supported in part by the National Research Centre for Asbestos Related Diseases, as part of the Strategic Award by the National Health and Medical Research Council of Australia. M.J.S. was supported by a National Health and Medical Research Council Senior Principal Research Fellowship and Program grant.
Abbreviations used in this paper: DC, dendritic cell; CBS, citrate-buffered saline; HA, hemagglutinin.