Hematopoietic progenitor kinase 1 (HPK1) is a hematopoietic cell-restricted member of the Ste20 kinases that acts as a negative regulator of T cell functions through the AP-1, NFAT, and NFκB pathways. Using HPK1-deficient (HPK1−/−) mice, we report in this study a novel role for HPK1 in dendritic cells (DCs). Specifically, we observed that matured HPK1−/− bone marrow-derived DCs (BMDCs) are superior to their wild-type (WT) counterpart in stimulating T cell proliferation in vivo and in vitro. Several characteristics of HPK1−/− BMDCs may account for this enhanced activity: Matured HPK1−/− BMDCs express higher levels of costimulatory molecules CD80, CD86, and I-Ab as well as produce more proinflammatory cytokines IL-12, IL-1β, TNF-α, and IL-6 than their WT littermates. The role of HPK1 as a proapoptotic molecule was assessed post activation with LPS, and results indicated that HPK1−/− BMDCs are significantly resistant to LPS-induced apoptosis. Our results led us to investigate the role of HPK1−/− BMDCs in tumor immunotherapy. Using a s.c. murine model of Lewis Lung Carcinoma, we found that HPK1−/− BMDCs eliminate established s.c. Lewis Lung Carcinoma more efficiently than their WT counterpart. Our data reveal a novel role for HPK1 as a negative regulator of DC functions, identifying its potential as a molecular target for DC-based immunotherapy against cancers.
Dendritic cells (DCs)4 are cellular bridges that couple the innate and adaptive immune systems. DC activation is a process that involves phenotypic and functional changes, both of which are orchestrated by a highly integrated network of molecular and biochemical signals that are initiated upon the binding of pattern recognition motifs to unique receptors on the surface of a DC. This interaction transitions DCs from an immature to a mature state and enables them to migrate from the periphery to the T cell-rich areas of regional lymph nodes and facilitate effector functions. This process relies on the up-regulation of receptors that are involved in lymphocyte activation and DC migration, such as the costimulatory molecules CD80 (B7.1) and CD86 (B7.2), MHC class II, and the chemotactic receptor CCR7 (1). Stimulation of T cell responses by activated DCs not only depends on the physical, coreceptor-mediated contact between a DC and a T cell, but is also governed by several other factors: The cytokine milieu that is generated by antigenic encounter (2), which can directly or indirectly affect the activation state of T cells and aid in a paracrine regulation of T cell-specific cytokines and their respective receptors, and factors such as chemokines and growth factors that are essential for the elimination of pathogens from the site of infection and the maintenance of protective immunity (3). Furthermore, both direct and indirect effects of DCs on the establishment of an effective immune response are highly dependent on the viability of DCs (4, 5). Numerous studies have implicated DC apoptosis in the persistence of infection and the promotion of carcinogenesis (5, 6, 7).
Hematopoietic progenitor kinase 1 (HPK1) is a serine/threonine kinase and a member of the germinal center kinase family (8, 9, 10). Via its ability to inhibit AP-1, NFAT, Erk, and NFκB-mediated gene transcription, HPK1 negatively regulates T and B cell responses (10, 11, 12, 13, 14). We have been interested in characterizing the biological role of HPK1 in T cells and have reported that HPK1 becomes catalytically activated upon stimulation by prostaglandin E2 in transformed hematopoietic cell lines (15, 16), while others have shown that HPK1 is also activated by another immunosuppressive molecule, TGF-β (17, 18). These findings expand the number of cell surface receptors that can activate HPK1 and suggested that HPK1 may serve as a negative regulator of receptor signal transduction in other hematopoietic cells. Although HPK1 is expressed in many differentiated hematopoietic cell types (19), its expression in DCs has never been identified.
In this report, we document for the first time, a function for HPK1 in murine DCs. Furthermore, we report enhanced DC maturation in the absence of HPK1, and demonstrate that HPK1−/− DCs acquire an enhanced ability to stimulate T cells when compared with wild-type (WT) DCs. This effect may be due to a combination of DC-specific factors that are enhanced in the absence of HPK1. Therefore, HPK1 appears to be a novel negative regulator of DC functions and may serve as a target of therapeutic interventions designed to enhance DC based immunotherapy.
Materials and Methods
C57BL/6 (H-2Kb), SJL/J (H-2Ks), and TCR-transgenic OT-1 (H-2Kb) mice were from Jackson ImmunoResearch Laboratories HPK1-deficient mice were generated and backcrossed to C57BL/6 background for nine generations (submitted for publication). HPK1−/− mice were healthy, reproduced with Mendelian ratios and had a normal life span (submitted for publication). All animal work has been performed in agreement with institutional Division of Laboratory Animal Resources regulations.
Abs, medium, and reagents
The following Abs were purchased from BD Pharmingen: Florescein isothyocyanate-conjugated hamster anti-mouse CD80, and CD11c, CD4, CD8, and R-PE-conjugated rat anti-mouse CD86, I-Ab, IL-2, and IFN-γ and H-2Kb and allophycocyanin-conjugated CD4. GolgiStop was also purchased from BD Pharmingen. RPMI 1640 (CellGro) supplemented with 10% Fetalplex (Gemini Bio-Products), 2–2-ME (2-ME, 50 μM) from Life Technologies, and l-glutamine (2 mM)/penicillin (100 U/ml)/streptomycin (100 μg/ml) from Gemini Bio-Products was used as a complete DC medium. Recombinant murine granulocyte and macrophage CSF was from R&D Systems. E. coli LPS, ConA, FITC-Dextran, and CFSE were from Sigma-Aldrich. All ELISA Abs and standards were purchased from R&D Systems.
Bone marrow-derived DC (BMDC) preparation
The femur and tibia were removed from mice and the bone marrow was flushed out using a 25-gauge needle. The resulting cell suspension was grown in complete DC medium plus 20 ng/ml recombinant murine granulocyte and macrophage CSF for 10 days in Petri dishes (2 × 106 cells/dish). FACS analyses were performed to determine the level of CD11c expression in differentiated bone marrow cells. Nonadherent cells were then transferred to tissue culture dishes (immature DCs), or were matured in the presence of 1 μg/ml LPS.
In vivo T cell stimulation
T cells were purified from the spleens of OT-1 mice by Lympholyte-M (Cedarlane Laboratories). Cells were resuspended at 5 × 106 cells in 100 μl of PBS, labeled with 5 μM CFSE and injected i.v. into the retro-orbital sinus of WT C57BL/6 or HPK1-deficient recipients. At the same time, either OVA (50 μg/ml) or PBS (as a control) was injected s.c. into the footpad of the same mice. For BMDC stimulation experiments, LPS-matured and OVA-loaded, or unpulsed WT or HPK1−/− BMDCs were injected at the same time as CFSE-labeled OT-1 T cells into WT recipients. After 48 h, the regional lymph nodes were removed and proliferation of OT-1 cells, as read by the extent of CFSE-dilution, was assessed using a FACSCalibur and analyzed by FlowJo software (Tree Star).
FACS and cytokine analysis
Immature and LPS-matured BMDCs were cultured in complete medium. Supernatants were collected after 24 h of stimulation for cytokine analysis or cells were resuspended in labeling medium (5% FBS, 0.1% sodium azide in 1 × PBS) for surface receptor staining. Cells (1 × 106) were incubated with the appropriate conjugated Abs for 50 min at 4°C. The cells are then washed twice with labeling medium and the pellet was resuspended in 1 × PBS and analyzed by FACS. For intracellular detection of cytokines, PMA (40 ng/ml), ionomycin (2 μM), and GolgiStop were added for 5 h to a 72 h MLR. Cells were first surface stained and then fixed with 2% paraformaldehyde, followed by permeabilization with a buffer containing 0.1% Saponin and 0.1% BSA in the presence of intracellular mAbs for 30 min at room temperature. For annexin V staining, cells were incubated with annexin V-PE for 15 min before analysis. Flow cytometry was performed on a FACSCalibur and analyzed with FlowJo software. Cytokine ELISA followed manufacturer’s protocol.
Mixed leukocyte reaction
Naive T cells from an SJL/J mouse spleen were purified by negative selection using a pan T cell isolation kit (Miltenyi Biotec). In a 96 round-bottom plate, purified naive T cells (2 × 105) were used as responders and incubated with varying dilutions of irradiated immature or 24 h LPS-matured HPK1−/− and WT BMDCs as stimulators. Anti-CD11c-coated magnetic beads (Miltenyi Biotec) were used to positively select splenic DCs from the spleens of either WT or HPK1−/− mice. Naive T cells plus 5 ng/ml ConA was used as a positive control for T cell stimulation. The background (DC proliferation) was subtracted from total proliferation and was found to be minimal. Cells were pulsed with 1 μCi of [3H]thymidine 18 h before harvest. Supernatant was collected before thymidine incorporation for IFN-γ and IL-2 ELISA.
Endocytosis assay of DCs with FITC-dextran
FITC-dextran (Sigma-Aldrich) was incubated at a final concentration of 0.5 mg/ml with immature BMDCs for 2 h at 37°C. BMDCs incubated with FITC-dextran at 0°C were used as a negative control for endocytosis. After incubation, cells were washed three times with ice-cold PBS, stained with PE-labeled anti-CD11c mAb, and analyzed by FACS.
Tumor model and vaccinations
Lewis Lung Carcinoma (LLC) cell line was obtained from American Type Culture Collection. Cells (1 × 105) were injected s.c. into the right flank of WT C57BL/6 mice and tumors were palpated every 2 days for the duration of the studies. DC vaccines were all prepared as follows: BMDCs were generated as above from either HPK1−/− or WT littermates. On day 10, immature BMDCs were pulsed for 5 h with LLC cell lysates generated by three rounds of freezing and thawing and 1 μg/ml LPS, or with LPS alone as a control. For the therapeutic model, BMDCs (2 × 106) were injected intratumorally into palpable tumors at 7 and 21 days post tumor challenge.
CTL and IFN-γ assay
Mice that were injected with WT or HPK1−/− Ag-loaded or immature BMDCs were sacrificed 21 days post tumor challenge. Regional lymphocytes from tumor-challenged animals were cultured with Irradiated LLC cells (20,000 rads) that were labeled with 200 μCi/ml 51Cr. Cytotoxicity against LLC was measured at variable effector to target (E:T) ratios after 6 h of coculture in V-bottom 96-well plates, in the presence of 10 U/ml IL-2. Percent lysis was calculated using the following formula: (% cytotoxicity = 100 × E− min/max − min) where “min” is the minimum (target cells alone) and “max” is maximum (100%) lysis. For IFN-γ analysis, lymphocytes from LLC-injected mice were stimulated ex vivo at a 40:1 ratio of T cells to irradiated LLC cells for 18 h, and IFN-γ levels were determined by ELISA.
Enhanced Ag presentation in HPK1-deficient mice
We hypothesized that HPK1, in addition to its known role as a negative regulator of cellular activation in T and B cells, also may perform a similar regulatory function in other hematopoietic cell lineages, including professional APCs such as DCs. Because the process of Ag presentation is required for T cell activation, we used the T cell response to OVA as a proxy reporter to quantitate the effectiveness of in vivo Ag presentation. OVA was injected into the footpad of WT and HPK1−/− mice and CFSE-labeled, OVA-specific T cells isolated from OT-1 TCR transgenic mice were transferred i.v. into these mice. Popliteal lymph nodes were removed 48 h after OVA injection and the CFSE-labeled OT-1 cells were analyzed for the dilution of the CFSE signals. When the number of OT-1 T cell divisions was inferred from CFSE-dilution data, it was clear that some OT-1 T cells had undergone at least seven cellular division in both WT and HPK1−/− hosts, as reflected by the number of distinct fluorescence peaks representing the decreasing strength of CFSE signals (Fig. 1,A). Over 90% of OT-1 cells recovered from the popliteal nodes of the HPK1−/− host had undergone at least four divisions, whereas slightly more than 10% of the OT-1 cells transferred to the WT host had proliferated to the same extent (Fig. 1 A). The finding that OT-1 T cells responded more robustly upon in vivo stimulation by APCs in HPK1−/− host was corroborated by the observation that more anti-OVA Vα2+ T cells clonally expanded in HPK1−/− hosts than in WT hosts (data not shown). As negative controls, OT-1 T cells injected into either WT or HPK1−/− hosts without OVA Ag did not proliferate beyond the baseline level in either host. Only 1.09% of OT-1 T cells proliferated in WT host and 2.2% of OT-1 proliferated in HPK1−/− host beyond the M1–3 gate, which encompasses cells that had undergone four or more cell divisions (supplemental Table I).5
HPK1−/− BMDCs are superior APCs
It is well documented that DCs play a crucial role in presenting OVA Ag to T cells. Thus, the enhanced T cell response to OVA observed in our previous experiments could be due to more efficient Ag presentation by HPK1−/− DCs. However, despite the known expression of HPK1 in hematopoietic cell types (10), the expression of HPK1 in DCs had not been documented. To assess this, we performed Western blot analysis on whole cell lysates prepared from BMDCs, using anti-HPK1 Ab as a probe. Analysis revealed that HPK1 is expressed in WT BMDCs and its absence was confirmed in HPK1−/− BMDCs (data not shown). To determine whether the loss of HPK1 in BMDCs would have any impact on their ability to stimulate T cells, CFSE-labeled OT-1 T cells were again used as an indicator of cellular proliferation. Forty-eight hours after CFSE-labeled OT-1 T cells were coinjected into C57BL/6 mice, with either OVA-pulsed WT or HPK1−/− BMDCs, they were harvested from popliteal lymph node cells and analyzed by FACS for the dilution of the CFSE fluorescence signal. We observed that >90% of OT-1 T cells had undergone at least four cell divisions when they were exposed in vivo to HPK1−/− BMDCs (Fig. 1 B), whereas only ∼30% of OT-1 T cells reached four cell divisions when WT BMDCs were used as APCs. To confirm the need for DC maturation and the presence of antigenic challenge, we found that immature WT BMDCs that were not pulsed with OVA, as well as immature and unloaded HPK1−/− BMDCs, failed to induce OT-1 proliferation in vivo. Only 2.8% of OT-1 cells that encountered WT BMDCs and 3.1% of OT-1 T cells that encountered HPK1−/− BMDCs had undergone more than four cell divisions (supplemental Table I). Because some OT-1 T cells stimulated by either WT or HPK1−/− BMDCs achieved seven to eight cellular divisions during this given time, the observed disparity in the percentage of cells must reflect the ability of HPK−/− BMDCs to directly prime T cells more efficiently than their WT counterpart. As a control, we found no difference in OT-1 T cell stimulation when mice were injected with unpulsed BMDCs from HPK1−/− or WT mice (supplemental Table I). Although we cannot rule out the possibility that HPK−/− BMDCs primed T cells more effectively by influencing the function of the endogenous APCs (e.g., via cytokine release in trans), we do not believe this is a likely possibility. Regardless of whether HPK−/− BMDCs achieved their superior T cell stimulation directly or indirectly, it is clear that OT-1 T cells responded more robustly to stimulation by transferred HPK−/− BMDCs in comparison to WT BMDCs.
To further characterize the effects that the loss of HPK1 incurs on DC function, we turned to an in vitro assay to determine the capacity of HPK1−/− vs WT BMDCs to stimulate T cells. Immature or LPS-matured BMDCs from HPK1−/− and WT mice were cocultured with varying numbers of allogeneic SJL/J splenic T cells. After 2 days of a MLR, wells containing mature HPK1−/− BMDCs already displayed greater T cell stimulation than ones stimulated with WT BMDCs (Fig. 2,A, left). In fact, after 2 days of an MLR, T cell stimulation by the HPK1−/− BMDCs had already achieved the maximal stimulation observed by WT BMDCs after 4 days of an MLR (Fig. 2,A, left and right). Similar enhanced MLR response was observed when SJL/J T cells were stimulated with HPK1−/− splenic DCs (supplemental Fig. 2). After 4 days, T cell stimulation by HPK1−/− BMDCs was nearly 2-fold greater than stimulation by WT BMDCs. When the T cell to BMDC ratio required to give equivalent T cell proliferation was compared, the number of HPK1−/− BMDCs needed was found to be nearly 16-fold less than the number of WT BMDCs needed (compare T:BMDC 1:0.0625 vs 1:1) (Fig. 2,A, left). Furthermore, intracellular cytokine analysis revealed that T cells produced more IL-2 and IFN-γ in response to stimulation by HPK1−/− BMDCs when compared with WT BMDC stimulation (Fig. 2, B–D, left and right). Analysis revealed that the enhanced IL-2 production in response to stimulation by HPK1−/− BMDC was attributable to the increase in the percentage of the IL-2-producing CD4+ T cells (42% KO vs 24% WT) (Fig. 2,E, left). Similar enhanced cytokine response was also observed in CD8+ T cells where the increased IFN-γ production is attributable to the increase in the percentage of the IFN-γ-producing CD8+ T cells (48% KO vs 18% WT) and, to a lesser extent, IFN-γ-producing CD4+ T cells (27% KO vs 18% WT) (Fig. 2 E, middle and right, respectively). These findings demonstrate, both in vitro and in vivo, that the lack of HPK1 renders BMDCs more potent APCs.
HPK1−/− BMDCs display higher expression of activation markers
The strength of T cell stimulation by DCs relies on a number of factors such as the up-regulation of costimulatory molecules and receptor-mediated endocytosis of pattern recognition motifs, all of which are acquired by a DC upon encounter with an activation stimulus. To test which, if any, of these factors contributed to the effect that we observed by HPK1−/− BMDCs, we determined whether the loss of HPK1 affected the expression levels of key DC coreceptors, which are up-regulated during the transition of immature DCs to a mature state. We began by examining the levels of CD11c on immature BMDCs, and observed no difference between WT and HPK1−/− BMDCs after 10 days of GM-CSF treatment (supplemental Fig. 1, A and B). This suggested that HPK1 does not affect the differentiation of bone marrow cells to BMDCs or the overall levels of CD11c on the surface of BMDCs. To test the effect of HPK1 on BMDC activation, cells were matured with LPS and stained with anti-CD80, anti-CD86, anti-I-Ab, and anti-H-2Kb Abs. After 24 h of incubation with LPS, BMDCs from HPK1−/− mice displayed a higher surface expression of CD80, CD86, I-Ab, and to a lesser extent H-2Kb when compared with those from their WT counterpart (Fig. 3 A). These data suggest that HPK1 may negatively regulate the LPS-induced up-regulation of DC activation markers, and that this may in part explain their superior T cell stimulation in the absence of HPK1.
Proinflammatory cytokine production is enhanced in HPK1−/− BMDCs
Upon maturation by LPS stimulation, DCs produce high levels of proinflammatory cytokines (20, 21), which are essential for DC-mediated activation of T cells. Because HPK1 is known to regulate cytokine production by T cells (22), we determined whether the lack of HPK1 would also affect cytokine production by BMDCs, which may have affected the magnitude of T cell stimulation by HPK1−/− BMDCs. Immature BMDCs were matured by LPS stimulation for 24 h and supernatants were collected for cytokine measurements by ELISA. There was a marked increase in IL-12, TNF-α, IL-6, and IL-1β production by LPS-matured HPK1−/− BMDCs as compared with WT BMDCs (Fig. 3 B). Not only is IL-12 necessary for an effective Th1 response, it also acts as an enhancer of CD8+ cytotoxic activity and subsequently induces the production of IFN-γ in T cells (23). Thus, these findings are consistent with the superior T cell stimulation by HPK1−/− BMDCs.
Normal endocytosis in HPK1−/− BMDCs
Immature BMDC are able to endocytose Ags with high efficiency using specialized receptors, such as FcR (24), mannose receptors (25), and phagocytic receptors (CD36) (26). This property enables them to internalize proteins, whole cell lysates, RNA, and apoptotic cells. This high endocytic capacity is lost during the maturation process, and is replaced by an up-regulation of MHC class II Ags whereby the processed Ags are presented. Thus, it was possible that the enhanced MHC class II expression and T cell stimulation was due to more efficient endocytosis by HPK1−/− BMDCs. To test endocytic capacity, immature WT or HPK1−/− BMDCs were incubated with FITC-Dextran for 2 h at 37°C, and CD11c+, FITC-positive cells were analyzed by FACS. Compared with controls incubated at a nonpermissible temperature (0°C), WT and HPK1−/− BMDCs incubated at permissive temperatures both displayed comparable levels of endocytosis (Fig. 3 C). This suggests that the lack of HPK1 did not enhance the endocytic capacity of BMDCs, and that the enhanced Ag presentation by HPK1−/− BMDCs is not due to superior endocytic capacity.
Resistance of HPK1−/− BMDCs to apoptosis
Because HPK1 has been shown to possess proapoptotic properties in T cells (24, 25), we assessed whether HPK1−/− BMDCs are resistant to apoptosis. Importantly, increased survival could explain some of the functional superiority of these BMDCs relative to their WT counterpart. Specifically, when BMDCs stimulate T cells in a coculture for over 48 h, it is possible that the enhanced T cell stimulation by HPK1−/− BMDCs is in part due to more viable BMDCs in the culture. Resistance to apoptosis may also partly explain the enhanced cytokine production. To test this possibility, BMDCs from WT or HPK1−/− animals were matured with LPS or left untreated (immature) for 24 to 120 h and cells were then stained with annexin V to measure the percent of BMDCs undergoing apoptosis. There was no difference in the apoptosis of immature cells, however, there was a clear difference in the annexin V+ population when mature WT and HPK1−/− BMDCs were compared. At 96 h, most of the WT BMDCs were annexin V+ (Fig. 3 D), whereas only ± 29% of BMDCs lacking HPK1 were positive for annexin V. This pattern remained consistent even at 120 h post LPS stimulation. Our results demonstrate that HPK1 is proapoptotic not only for T cells but also for DCs, and that this may account for some of the enhanced capacity in Ag presentation by BMDCs lacking HPK1.
Vaccination of tumor-bearing mice with HPK1−/− BMDCs results in complete tumor eradication
The absence of HPK1 resulted in enhanced Ag presentation and BMDC survival; therefore, we assessed whether HPK1−/− BMDCs might make more potent antitumor vaccines. To test the therapeutic efficacy of HPK1−/− BMDCs as tumor vaccines, we injected 1 × 105 LLC cells s.c. into the right flank of WT mice. By day 5, tumors were palpable, and on day 7 and 21, 2 × 106 tumor lysate-pulsed, LPS-matured WT or HPK1−/− BMDCs were injected intratumorally into WT C57BL/6 tumor-bearing mice. Although vaccination with tumor lysate-loaded WT BMDCs marginally reduced tumor growth when compared with unloaded controls, HPK1−/− tumor lysate-loaded BMDCs completely eliminated tumor growth (Fig. 4,A). It was also clear that immature, unloaded HPK1−/− BMDCs were also able to reduce tumor growth to the level achieved by lysate-loaded WT BMDCs. This is consistent with the in vitro results and is, therefore, likely to be due to a more efficient activation of immature HPK1−/− BMDCs by tumor Ags. In agreement with our hypothesis, BMDCs lacking HPK1 were effective in the eradication of established tumors, while mice injected with tumor-pulsed WT BMDCs were only moderately effective (Fig. 4 A). In fact, although tumor-pulsed WT BMDCs slowed disease progression, tumor growth reached maximum size allowed institutionally by day 50, and all mice examined were moribund. However, mice injected with tumor-pulsed HPK1−/− BMDCs remained tumor-free up to 60 days post tumor challenge (data not shown).
HPK1−/− BMDCs are more potent stimulators of CTL response against LLC
The success of BMDCs as tumor vaccines relies on their ability to stimulate an effective T cell response against the tumor. We have already established that HPK1-deficient BMDCs are more efficient APCs in vivo; therefore, we assessed whether tumor elimination in mice injected with HPK1−/− BMDCs may be due to the enhanced activation of a CTL response against the tumor. To compare antitumor T cell responses between tumor-bearing mice injected with WT or HPK1−/− BMDCs, cells from the draining lymph nodes of mice that have been injected with activated tumor-loaded or unloaded BMDCs were collected 14 days post LLC injection. Irradiated and 51Cr-labeled LLC cells were cocultured in the presence of draining inguinal lymphocytes at varying effector to target (E:T) ratios, and the ability to lyse LLC cells was determined ex vivo. When lymphocytes from tumor-bearing mice that have been injected with activated HPK1−/− tumor loaded BMDCs were cocultured with irradiated LLC cells at varying E:T ratios, they were four times more effective at killing LLC cells than those from mice that have been vaccinated with tumor loaded WT BMDCs (Fig. 4,B). Furthermore, when supernatants from this assay were collected and analyzed for the production of IFN-γ by ELISA, significantly more IFN-γ was detected in the supernatants collected from cultures containing T cells isolated from mice vaccinated with HPK1−/− BMDCs vs cultures containing T cells from mice vaccinated with WT BMDCs (Fig. 4 C). The increased potency of T cells from HPK1−/− BMDC vaccinated animals in killing LLC supports the proposal that tumor rejection in these mice is due to better Ag presentation, suggesting that HPK1-deficient BMDCs may aid in the development of more successful antitumor vaccines.
Catalyzed by the pioneering work of Inaba, Steinman, and colleagues (27) which demonstrated the ability of murine BMDCs to be cultured ex vivo from bone marrow precursors, DCs have become an important therapeutic tool for tumor immunologists for the past decade. By comparing immature and mature DCs, studies have concluded that maturation is required for effective immunity in cancer patients (28, 29). Therefore, it is no surprise that the most effective DC vaccines rely on one or several approaches to enhance DC maturation/activation.
The ability of DCs to fine-tune their function is necessary for efficient regulation of the immune response and, therefore, requires a careful balance between both positive and negative signals. Despite a significant gain in our understanding of the positive regulation by cytokine and TLR signals, our knowledge of negative regulation of DC functions remains rudimentary. In this study, we have identified HPK1 as a novel negative regulator of Ag presentation. Specifically, BMDCs lacking HPK1 stimulated T cells more efficiently, and we have demonstrated that this may in part be due to the enhanced survival of such DCs, which in turn affects their expression of costimulatory molecules and production of proinflammatory cytokines, ultimately leading to a more effective tumor vaccine. Thus, the role of HPK1 in the negative regulation of DC functions may act as a break on Ag presentation and, therefore, substantially effects the regulation of T cell activation.
There are two issues to consider in understanding how the loss of HPK1 enhances BMDC function: The enhanced process of maturation and the resistance to apoptosis. Although it is clear that the lifespan of DCs affects the efficiency of T cell responses, most studies have dealt with apoptosis and DC maturation as separate entities. The NF-κB and JNK pathways have been implicated in the regulation of DC survival and maturation, and biochemical studies have attributed a role for HPK1 in both pathways (10, 12, 22, 30, 31, 32, 33, 34, 35). Recent studies have implicated prosurvival pathways with enhanced DC maturation and T cell stimulation (5, 30, 36, 37); therefore, it is likely that both effects play a role in the phenotype that we observe in HPK1−/− BMDCs. Additionally, while both Bcl-2 and Bcl-XL have been implicated in the prosurvival pathway of Toll-like-receptor-stimulated DCs (4, 5), a recent study has demonstrated the involvement of Akt in the anti-apoptotic properties and subsequent function of DCs via Bcl-2 (5). Not only does Akt block signaling to Bcl-2, but it is also able to regulate the NFκB pathway. Using overexpression studies in T cell lines, both pathways are shown to be affected by HPK1 (12, 38, 39). We are currently deciphering the pathways that HPK1 engages which may regulate the functional outcome of DC maturation vs apoptosis, in anticipation of bridging the connection between DC survival and function. These will be important experiments to pursue because HPK1 appears to be a negative regulator of DCs and, therefore, an important parameter to manipulate for the treatment of established tumors, and maybe other immune-involved diseases.
DCs have recently become a major target for tumor immunotherapy because of their potency in stimulating T cells. As elements of the innate and adaptive immune system work in concert to eliminate malignant cells, the efficiency of tumor-specific T cells certainly relies on effective Ag presentation. Although immature or semimature DCs may lead to tolerance, Ag-presentation by DCs can be modulated in certain physiologic and pathologic situations, and potential negative regulatory roles have been suggested for subpopulations of DCs that have been exposed to immunosuppressive factors that are secreted by tumors such as TGF-β and IL-10 (40, 41, 42, 43). Allowing DCs to process tumor cell lysates and present them to T cells has been proposed as a powerful tool for generating tumor-specific T cell responses. This approach, however, relies on a multitude of factors such as tumor heterogeneity, and the efficiency of Ag presentation, maturation status, and longevity of DCs. Therefore, there are more interventions that may be available to allow DC-based therapies to more effectively eradicate established disease.
The absence of HPK1 on BMDCs resulted in the eradication of established tumors and an effective T cell stimulation that is certainly, but perhaps not exclusively, dependent on higher expression of costimulatory molecules, proinflammatory cytokine production, and survival. The fact that HPK1 is a negative regulator of BMDCs represents a novel therapeutic target for a multitude of DC functions. Careful consideration of the possible autoimmune and inflammatory consequences of knocking down a negative regulator is necessary, and a combination of therapies may be necessary to generate successful antitumor vaccines and maintain systemic homeostasis.
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.
Abbreviations used in this paper: DC, dendritic cell; HPK1, hematopoietic progenitor kinase 1; WT, wild type; BMDC, bone marrow-derived DC; LLC, Lewis Lung Carcinoma.
The online version of this article contains supplementary material.