Abstract
Acute myeloid leukemia (AML) is an aggressive hematologic malignancy with a dismal prognosis. Immunotherapeutic approaches using single agent checkpoint inhibitors have thus far shown limited success. We hypothesized that successful adaptive anti-AML specific immune responses require additional modulation of innate immunity. DMXAA exposure resulted in modest apoptosis of C1498 AML cells with a subtle increase in PD-L1 expression and limited production of IL-6 and IFN-β. In contrast, DMXAA + anti-PD-1 ab, but not either agent alone, significantly decreased in vivo disease burden and prolonged overall survival in C1498 engrafted leukemic mice. Combination-treated mice demonstrated increased memory T-cells and mature dendritic cells, lower numbers of regulatory T-cells and evidence of leukemia apoptosis. Furthermore, these effects were associated with markedly increased serum levels of type I interferon and interferon gamma. We demonstrate that combining an innate immune agonist with a checkpoint inhibitor synergistically improved anti-tumor activity in a preclinical AML model.
Keywords: immunotherapy, AML, PD-1, innate immune agonist, checkpoint inhibitor
Introduction
Acute myeloid leukemia (AML) is an aggressive hematological malignancy arising from an immature myeloid progenitor. High dose chemotherapy, commonly known as 7+3, has been the standard of care for AML for many years but is associated with significant mortality and morbidity, particularly in older individuals who constitute the majority of cases. Although novel immunotherapeutic combination approaches appear to offer clinical activity, these data require validation in a phase 3 study [1,2]. Present approaches offer long-term survival in only 20–30% of patients [3], highlighting the need for novel therapeutic approaches.
A critical mechanism of immune escape employed by AML cells within the microenvironment is to exploit the altered interaction with PD-1, an immune checkpoint receptor expressed on T-cells [4]. The ligands for PD-1 (PD-L1 and PD-L2) are typically expressed on the surface of antigen presenting cells. Binding of these ligands to PD-1 triggers inhibitory signaling in T-cells, resulting in decreased cytokine production and suppressed T-cell proliferation. Both PD-L1 and PD-L2 expression has been shown to be upregulated greater than 2-fold on leukemic cells as compared to controls, which have been associated with poor prognosis [5,6]. Prior experiments demonstrated that up-regulation of PD-L1 and PD-L2 on a human AML (HL60) cell line led to severe suppression of T helper (Th)-cell responses. Targeting PD-1 and PD-L1 in an immunocompetent murine AML model reduced tumor burden and increased the proliferation and function of cytotoxic T-cells at tumor sites with improved long-term survival [7,8]. However, clinical use of targeting checkpoints alone as a means of immunotherapy has been met with limited success [1,2].
An attractive agent to consider for the augmentation of checkpoint inhibition is 5,6-dimethylxanthenone-4-acetic acid (DMXAA), which not only disrupts tumor vasculature [9], but also activates the stimulator of interferon (IFN) genes (STING) pathway [10]. In two AML models, DMXAA has been found to induce type I IFN signaling as well as TNF-α and IL-6, promoting dendritic cell maturation. This results in a striking maturation of leukemia-specific T-cells, which leads to prolonged overall survival in leukemic mice. Here we demonstrate that combination therapy with PD-1 inhibition and the innate immune agonist and anti-vascular agent DMXAA reduces disease burden and improves overall survival via enhancement of the innate immune system and modulation of the adaptive response.
Materials and Methods
AML cell line
C1498 is a murine AML cell line that spontaneously developed in C57BL/6 mice. A stable transfectant of this cell line, C1498FFDsR (a generous gift from Dr. Bruce Blazar, Minneapolis, MN) was utilized in these experiments. This transfected cell line expresses fluorescent Discoma coral-derived protein DsRed2 and firefly luciferase [11].
Cell culture
C1498FFDsR cells were incubated in Dulbecco modified Eagle complete media (DMEM-c) at 35 degrees. Cells were exposed to increasing concentrations of DMXAA (Selleck Chemicals) ranging from 1–1000 μg/ml to assess PD-L1 expression and apoptosis. Murine interferon gamma (EMD Millipore) served as a positive control for PD-L1 expression, and matrine (Selleck Chemicals) served as a positive control for apoptosis.
Flow cytometry
For in vitro experiments assessing PD-L1, C1498FFDsR cells were washed with PBS then surface labeled with PD-L1-APC (BioLegend). Apoptosis was assessed in vitro with C1498 cells employing the FITC Annexin V apoptosis kit (BioLegend). Antibodies utilized for in vivo analyses were obtained from BioLegend, BD Horizon, and Thermo Fisher. For both in vitro and in vivo experiments, events were then collected using the FACS LSRII (BD Bioscience), and were gated on 15,000 events. FlowJo v10 was utilized to perform data analysis.
Cytokines
For in vitro assays C1498FFDsR cells were treated with DMXAA or PBS and cell supernatant was obtained 48 hours later. ELISA was performed for IFN-β and IL-6 using pre-coated plates (BioLegend) according to the manufacturer’s protocol. For in vivo assays, C1498FFDsR inoculated mice were harvested 24 hours (day 21 post inoculation) after receiving either 1 dose of vehicle, DMXAA, anti-PD-1 or DMXAA and anti-PD-1. Luminex assays to assess for Interferon- α (IFN-α), IFN-β, IFN-γ, IL-6, and TNF-α were performed by the Roswell Park Flow Cytometry Core utilizing Affymetrix pre-coated plates.
Immunohistochemistry
VEGFR3
Formalin-fixed paraffin sections were cut at 4μm, placed on charged slides, and dried at 60°C for one hour. Slides were cooled to room temperature and added to the Dako Omnis autostainer, where they were deparaffinized with Clearify (American Mastertech; catalog #CACLEGAL ) and rinsed in water. Flex TRS High (Dako; catalog #GV804) was used for target retrieval for 15 minutes. Slides were incubated with VEGFR3 antibody (BD Pharmigen; catalog# 552857) for 30 minutes at 1/40 (3.13 IgG). Rabbit Anti-Rat (Agilent; E0468) which was applied for 30 minutes followed by Rabbit Envision (Agilent K4003) for 30 minutes. DAB (Diaminobenzidine) (Dako; catalog #K3468) was applied for 5 minutes for visualization. Slides were counterstained with Hematoxylin for 8 minutes then put into water. After removing slides from the Omnis they are dehydrated, cleared and coverslipped.
Caspase 3
Formalin-fixed paraffin sections were cut at 4μm, placed on charged slides, and dried at 60°C for one hour. Slides were cooled to room temperature and added to the Dako Omnis autostainer, where they were deparaffinized with Clearify (American Mastertech; catalog #CACLEGAL ) and rinsed in water. Flex TRS High (Dako; catalog #GV804) was used for target retrieval for 30 minutes. Slides were incubated with Caspase 3 antibody (Cell Signaling, 9661) for 30 minutes at 1/50 (1.26 IgG). Rabbit Linker (Dako, GV809) was applied for 10 mins followed by Rabbit Envision (Dako #K4003) for 30 minutes. DAB (Diaminobenzidine) (Dako; catalog #K3468) was applied for 5 minutes for visualization. Slides were counterstained with Hematoxylin for 8 minutes then placed into water. After removing slides from the Omnis they are dehydrated, cleared and coverslipped.
CD163
Formalin and zinc fixed paraffin sections were cut at 4μm, placed on charged slides, and dried at 60°C for one hour. Slides were cooled to room temperature and added to the Dako Omnis autostainer, where they were deparaffinized with Clearify (American Mastertech; catalog #CACLEGAL ) and rinsed in water. Flex TRS High (Dako; catalog #GV804) was used for target retrieval for 30 minutes. Slides were incubated with CD163 antibody (Abcam, ab182422) for 30 minutes at 1/300 (2.24 IgG). Rabbit Envision (Dako #K4003) was applied for 30 minutes. DAB (Diaminobenzidine) (Dako; catalog #K3468) was applied for 5 minutes for visualization. Slides were counterstained with Hematoxylin for 8 minutes then placed into water. After removing slides from the Omnis they are dehydrated, cleared and coverslipped.
Mice
C57BL/6 mice, 6 weeks old at study, were obtained from The Jackson Laboratory and were maintained in the animal facility at Roswell Park Cancer Institute (RPCI). All mice were housed in micro-isolator cages under specific pathogen-free conditions. All of the experiments were performed under the RPCI Institutional Animal Care and Use Committee protocol.
In vivo AML model
All animal experiments were conducted in accordance with standard institute animal care and use committee protocols. C57BL/6 mice obtained from Jackson Laboratory were injected through the tail vein with C1498FFDsR (5 × 106) to achieve systemic engraftment to the degree reported in the literature [8,11,12]. Seven days following AML injection, they were initiated on treatment with vehicle, DMXAA, anti-PD-1 ab (Bio Xcell), or combination of DMXAA and anti-PD-1 ab. Treatment with either DMXAA at 20mg/kg or vehicle was administered intraperitoneally every four days for a total of six weeks. Anti-PD-1 ab was administered intraperitoneally at 10mg/kg on days 7, 10, 13, and 16.
Bioluminescent Imaging
Bioluminescent imaging with a xenogen IVIS imaging system assessed engraftment. Firefly luciferin substrate (0.05mL; 15mg/mL) was injected intraperitoneally and the IVIS imaging was performed sixty seconds after substrate injection. Data was analyzed and presented as photon counts per area.
Statistical Analysis
Survival curve was calculated using the Kaplan-Meier product-limit method. Differences between groups in the survival study were determined using log-rank statistics. For all other data, a Student t test was used to analyze differences between groups, and results were considered significant if the P value was less than or equal to 0.05.
Results
DMXAA exerts in vitro effects on murine AML cells
It has been previously reported that treatment with DMXAA exerts a myriad of effects including vascular disruption and endothelial apoptosis [13]. Here, we demonstrate that DMXAA exerts modest anti-leukemic activity. DMXAA (concentrations ranging from 1–1000 μg/ml) induced apoptosis of the murine AML cell line C1498 in vitro at 24 hours (24h) in a dose dependent manner (Figure 1A), although concentrations resulting in significant apoptosis were higher than physiologically achievable. DMXAA functions as a STING agonist [10], we aimed to characterize the immunomodulatory properties of this compound in our AML model, including effects on PD-L1 expression and cytokine production. Exposure of C1498 cells in culture to escalating doses of DMXAA (1–100μg/ml) or IFN-gamma (positive control) induced a subtle increase in the level of PD-L1 expression on these AML cells (Figure 1B). Furthermore, C1498 cells exposed to higher doses of DMXAA (10–100μg/ml) for 48h were able to generate measurably higher levels of IL-6 and (IFN-β) expression in cell supernatants (Figure 1C-D).
Figure 1. In vitro cytotoxic and immunomodulatory effects of DMXAA on murine AML cells.
(A) C1498 cells exhibit a dose dependent apoptotic cell death upon exposure to increasing concentrations of DMXAA at 48h. (B) Exposure of C1498FFDsR cells to DMXAA for 48h results in an increase in PD-L1 expression, as well as increased production of IL-6 (C) and INF-β (D). * p < 0.05. Error bars represent pooled data from at least 2 independent experiments and are represented as mean ± SEM.
Combined DMXAA and anti-PD-1 antibody treatment decreases AML burden and prolongs overall survival in an immunocompetent mouse model
Following completion of our in vitro work, we sought to determine whether our innate immune agonist might enhance the immunologic response to PD-1 blockade in vivo. Here, we evaluated the impact of DMXAA treatment in an immunocompetent murine model of AML [14]. Immunocompetent C57BL/6 mice were inoculated with stably transfected C1498 murine AML cells expressing luciferase and the fluorescent protein DSRed2 (5 × 106 cells/mouse) [7,8,14,15]. On day 7 following inoculation, animals were treated with vehicle, DMXAA (400μg intraperitoneally every 4 days x 6 weeks) anti-murine PD-1 ab (200μg intraperitoneally every 3 days x 4 doses) or the combination of DMXAA + anti-PD-1 antibody.
Treatment was well tolerated, without any systemic evidence of toxicity in treated animals. Treated animals and controls were followed for overall survival and total body bioluminescent imaging was performed on a weekly basis to measure disease burden. We observed that combination therapy with DMXAA + anti-PD-1 ab resulted in a marked decrease in disease burden (measured by total body bioluminescence) compared with either treatment as a single agent or vehicle controls (p<0.05) (Figure 2A,B). Time to morbidity requiring euthanasia (by pre-specified criteria) was significantly longer in the treatment groups compared with controls: vehicle = 29 days, DMXAA = 30 days, anti-PD-1 ab = 39 days, and combination DMXAA + anti-PD-1 ab = 49 days (Figure 2C). Combination therapy resulted in longer overall survival than any single agent therapy (DMXAA vs. DMXAA+anti-PD-1 ab, (p=0.0002); anti-PD1 ab vs. DMXAA+anti-PD-1 ab, (p=0.0451) (n=8–10 mice per group).
Figure 2. Combination DMXAA and anti-PD-1 antibody therapy reduces systemic leukemic burden and prolongation of overall survival in a preclinical AML mouse model.
On day 7 post inoculation with C1498FFDsR cells, animals were treated with vehicle control, DMXAA, anti-PD-1 ab or DMXAA + anti-PD-1 ab. A. Images were taken at day 26 post inoculation, and are representative of 13–16 animals/group. B. Bioluminescence of pooled experiments at day 26 post inoculation. C. Following treatment with either vehicle control, DMXAA, anti-PD-1 ab or DMXAA + anti-PD-1 ab animals were followed for overall survival. * p < 0.05. Error bars represent pooled data from at least 2 independent experiments and are represented as mean ±SEM. D.
Utilizing immunohistochemistry for caspase 3, we showed increased apoptosis in leukemia bearing animals receiving combination therapy with DMXAA and anti-PD-1 ab (Figure 3). Based upon these findings, we hypothesized that our combination therapy had enhanced the adaptive immune response to AML, increasing the infiltration of inflammatory cells to sites of leukemic burden which induced apoptosis of the leukemia cells.
Figure 3. DMXAA + anti-PD-1 therapy results in increased leukemic apoptosis in vivo.
20 days following treatment with either vehicle control, DMXAA, anti-PD-1 ab or DMXAA + anti-PD-1 ab animals were harvested. Caspase 3 positive leukemic cells as identified through H&E were found to be present at increased frequency in the liver of animals treated with combination therapy (D), compared to single agent DMXAA (B). anti-PD-1 ab (C), or vehicle control (A).
Combination DMXAA + anti-PD-1 antibody therapy induces in vivo immunomodulatory effects
Given our observation that combination therapy produced improved survival and induced apoptosis in our immunocompetent model of AML, we sought to characterize the mechanism for synergy. We hypothesized that co-administration of innate and adaptive immune agonists resulted in an enhanced anti-tumor inflammatory response. Specifically, we anticipated increased infiltration of tumor sites with CD44+ T-cells and mature dendritic cells due to enhanced local production of inflammatory cytokines [16–19]. We utilized our immunocompetent model to assess these hypotheses. Following inoculation with C1498, animals were treated with either vehicle, DMXAA, anti-PD-1 ab, or combination therapy. Animals were euthanized 20 days following therapy (before they demonstrated morbidity under any of our conditions) and sites of disease engraftment including bone marrow, spleen, liver and serum were harvested for testing.
We observed markedly elevated levels of cytokines in the serum of combination treated mice, including type I IFNs (IFN-α, IFN-β), IL-6, TNF-α, and IFN-γ compared to control animals treated with either single agent or vehicle (Table 1). These cytokines are the result of increased STING activity through the NF-κB and STAT6 pathways and indicate the function of CD44+ T-cells. Furthermore, we confirmed a shift in activated CD44+ T cells (CD3+, CD8+, CD44+) with enhanced infiltration of this population into sites of disease including the bone marrow, spleen and liver of leukemic mice (Figure 4). A larger proportion of mature dendritic cells were also observed in the bone marrow of animals receiving both single agent DMXAA and combination therapy (Figure 5) compared to control and single agent anti-PD-1 ab, although no differences were appreciated in the liver or spleen of these animals (data not shown). Animals receiving DMXAA also demonstrated evidence of vascular disruption compared to vehicle treated animals as determined by vessel diameter (Supplementary Fig. S1), which might account for the restriction of dendritic cell maturation to the bone marrow compartment. No changes in angiogenesis were seen in the spleen or liver of leukemic mice. Our laboratory has previously shown that DMXAA increases the delivery of subsequently administered chemotherapy to the bone marrow presumably due to these changes in vascularity [20].
Table 1.
Effects of DMXAA + anti-PD-1 therapy on cytokine levels in vivo
| Group | IFN-α | IFN-β | IFN-γ | IL-6 | IP-10 | TNF-α |
|---|---|---|---|---|---|---|
| Control | ||||||
| Mean (pg/mL) | 8.45 | 0 | 0.36 | 1.30 | 7.60 | 0.18 |
| SD | 11.95 | 0 | 0.12 | 0.49 | 0.06 | 0.25 |
| Vehicle | ||||||
| Mean (pg/mL) | 28.55 | 2.10 | 1.43 | 29.41 | 25.97 | 4.68 |
| SD | 10.54 | 1.88 | 0.43 | 16.69 | 10.31 | 3.31 |
| DMXAA | ||||||
| Mean (pg/mL) | 38.93 | 12.39 * | 8.84 | 183.67 * | 65.40 * | 13.60 * |
| SD | 17.61 | 3.99 | 4.04 | 21.76 | 9.17 | 7.37 |
| PD-1 | ||||||
| Mean (pg/mL) | 337.93 | 135.81 | 316.31 | 8791.22 | 63.27 | 27.40 |
| SD | 482.13 | 195.17 | 486.25 | 13601.31 | 64.77 | 35.96 |
| DMXAA/PD-1 | ||||||
| Mean (pg/mL) | 3240.00 ** | 688.50 ** | 2291.67 ** | 33900.00 ** | 149.17 ** | 523.00 ** |
| SD | 737.37 | 197.82 | 1354.20 | 6024.28 | 29.03 | 303.23 |
denotes P < 0.05 compared to vehicle.
denotes P < 0.05 compared to DMXAA
Figure 4. DMXAA + anti-PD-1 therapy is associated with enhanced CD44+ T-cells within various disease sites.
Samples were harvested from leukemia engrafted mice 20 days following treatment with vehicle control, DMXAA, anti-PD-1 ab or DMXAA + anti-PD-1 ab. Mean fluorescence intensity (MFI) was increased consistent with increased numbers of (CD3+, CD8a+, CD44+) T-cells within the (A) bone marrow, (B), spleen, and (C) liver. Representative of 3 independent experiments.
Figure 5. DMXAA + anti-PD-1 therapy is associated with increased numbers of mature dendritic cells in the bone marrow.
Samples were harvested from leukemia engrafted animals 20 days following treatment with vehicle control, DMXAA, anti-PD-1 antibody or DMXAA + anti-PD-1 antibody. Representative plots demonstrate increases in (A) CD11c+ CD86+, (B) CD11c+ CD80+ and (C) CD11c MHCII+ dendritic cells in the bone marrow of treated mice. Representative of 3 independent experiments.
A trend towards lower immunosuppressive regulatory T-cell numbers was observed in combination treated animals, which did not achieve statistical significance (Supplementary Fig. S2). We identified more M2 CD163+ macrophages in the spleens of leukemic mice treated with both DMXAA and anti-PD-1 ab (Supplementary Fig. S3D) compared to single agent and vehicle treated controls (Supplementary Fig. S3A-C). Given their established roles in T-cell suppression and immune evasion, we assessed our mice for the presence of polymorphonuclear myeloid-derived suppressor cells ((PMN)MDSCs), and monocytic myeloid-derived suppressor cells (mMDSCs) within the microenvironment. While these populations were identified in all our mice, we did not appreciate any differences in treated compared with control animals (Supplementary Fig. S4). Finally, expression of PD-L1 and PD-L2 on leukemic blasts harvested from the bone marrow, spleen and liver did not change with treatment (Supplementary Fig. S5).
Discussion
Our work highlights the important, interdependent relationship between the innate and adaptive immune system. Previous studies have shown this approach to be active in solid tumors, here we show a similar effect in a historically immunologically “cold” tumor and demonstrate that combined therapy targeting the innate and adaptive immune system can produce significant anti-tumor activity in a preclinical immunocompetent model of AML. Using this model, we examined the tumor microenvironment to better understand the impact of our combination on cytokine signaling and immune cell infiltrates to help explain our observations.
Type I interferon responses have been demonstrated to be critical for endogenous priming of T-cell against tumors [21,22]. A significant deficiency in such responses has been shown in AML [10]. Gene expression signatures suggesting an active Type I interferon response in association with active CD8+ T-cells has been seen in patients with melanoma [23]. These signatures are thought to result from dendritic cell-mediated recruitment of cytotoxic CD8+ tumor infiltrating lymphocyte (TIL) to the tumor microenvironment [24,25]. We and others have identified a deficiency in type I interferons in patients with and mouse models of AML. We and others have shown that the addition of a STING agonist to leukemic blasts in vitro results in the production of IL-6 and IFN-β and increased expression of PD-L1. These findings provide the rationale for combination therapy with a STING agonists and a checkpoint inhibitor to enhance the activity of an anti-AML immune response.
We here confirmed prior results demonstrating that STING agonists produce tumor apoptosis in AML blasts treated in vitro [24,26]. We found that in our hands, induction of apoptosis required higher than physiologically achievable concentrations of DMXAA, suggesting that direct induction of apoptosis is unlikely to be the primary mechanism of this agent. In vitro, our STING agonist either alone or in combination with our checkpoint inhibitor produced low-level induction of PD-L1, IFN-β and IL-6. By contrast with these findings, immunocompetent leukemia bearing mice treated with DMXAA+ anti-PD-1 ab produced a marked increase in the serum levels of IFN-β and IL-6. Combination treated mice lived longer than control mice or those treated with either agent alone, and we observed immune-cell infiltration at sites of AML disease burden. These findings suggests 1) Combinations of DMXAA +anti-PD-1 exert their activity through interactions with the intact immune microenvironment in AML-bearing animals 2) such a combination approach might enhance immune recognition of AML cells.
Here we demonstrate that robust immune mediated anti-tumor activity can occur even in an immunologically “cold” tumor like AML. We observe that STING signaling can be restored in such tumors through the addition of DMXAA, increasing production of a type 1 IFN response. along with augmented IL-6 and TNF-α due to downstream NF-κB and STAT6 signaling. Increased production of IFN-γ was seen in our treated mice, supporting the hypothesis that tumor killing was mediated by the activity of CD44+ T-cells. These cytokines were markedly elevated in animals receiving combination therapy compared to single agent and vehicle treated controls. We observed increased apoptosis with combination therapy and hypothesize that this results in increased circulating tumor-specific antigens which can feedback to promote further STING signaling.
We showed that in our model DMXAA disrupts vasculature in the bone marrow microenvironment, but we do not believe that this mechanism is causal for improved survival. Previous investigators have combined bevacizumab with standard chemotherapy for patients with AML, without any improvements in complete remission rates or event free survival [27]. We hypothesize that STING agonism is the primary mechanism of action for this agent (in combination with our anti-PD-1 ab). We believe this combination enhances immune recognition of AML cells and is responsible for the improved survival outcome in our model. Our laboratory previously demonstrated that DMXAA induced vascular disruptions increases chemotherapy delivery to the microenvironment [20]. We hypothesize that this may enhance delivery of DMXAA itself to the bone marrow microenvironment and thereby facilitate our observation of improved maturation of bone marrow resident dendritic cells.
Therapeutic resistance to immune therapies is driven by immunosuppressive populations. Surprisingly, in our model we did not observe changes in number of bone marrow or spleen regulatory T-cells or MDSC populations. We did identify an increase in the number of splenic CD163+ pro-tumor M2 macrophages in combination therapy animals, a population of cells which might provide resistance to immunotherapy. Critically, combination therapy was effective in our model irrespective of PD-L1 status and blasts harvested following in vivo treatment showed no changes in expression of PD-L1. This is in contrast to the findings from other groups which suggest that in other tumors, PD-L1 expression decreases following anti-PD-1 therapy; these changes do not appear to correlate with therapeutic response [28].
DMXAA preferentially targets the murine STING pathway and our data suggest that an intact immune microenvironment is required for effective activity [29]. Our work was isolated to a single model due to the challenge of assessing T-cell mediated responses in other model systems [7,8,11,12,15]. While DMXAA itself is not immediately translatable to a clinical trial, human STING agonists are in pre-clinical development. Our work suggests that such agents should be rationally combined with checkpoint inhibition for patients with AML, but it also demonstrates that the immunological milieu is likely to be a critical component of any activity [30]. Our data support the potential for immune therapies in diseases like AML but also highlight the challenges of such approaches, which require adequate function of both adaptive and cellular compartments of the immune system. Such strategies will depend upon the development of sophisticated understanding of how bone marrow diseases impact the immune environment.
Supplementary Material
Supplemental Figure S1. DMXAA results in vascular disruption within the bone marrow. Samples were harvested from leukemia engrafted animals 20 days following treatment with vehicle control or single agent DMXAA. Bone marrow sections were stained with VEGFR. Representative photomicrographs demonstrate increased vascular disruption as measured by increase in vessel diameter. Error bars represent pooled data from at least 2 independent experiments and are represented as mean ± SEM.
Supplemental Figure S2. DMXAA + anti-PD-1 therapy does not impact regulatory T-cell populations. Samples were harvested from leukemia engrafted animals 20 days following treatment with vehicle control, DMXAA, anti-PD-1 antibody or DMXAA + anti-PD-1 antibody. A trend towards decreased regulatory T-cells (CD4+, CD25+, FOXP3+) was not observed in the (A) bone marrow but was observed in the (B) spleen and (C) liver of leukemic mice. NS = not statistically significant. Error bars represent pooled data from at least 2 independent experiments and are represented as mean ± SEM.
Supplemental Figure S3. DMXAA + anti-PD-1 therapy recruits more CD163+ macrophages. Samples were harvested from leukemia engrafted animals 20 days following treatment with vehicle control, DMXAA, anti-PD-1 antibody or DMXAA + anti-PD-1 antibody. Representative photomicrographs demonstrate CD163+ macrophages were found to be present at increased frequency in the spleen of animals treated with combination therapy (D), compared to single agent anti-PD-1 ab (C), DMXAA (B) or vehicle control (A).
Supplemental Figure S4. DMXAA + anti-PD-1 therapy does not affect populations of myeloid-derived suppressor cells. Samples were harvested from leukemia engrafted animals 20 days following treatment with vehicle control, DMXAA, anti-PD-1 antibody or DMXAA + anti-PD-1 antibody. Treatment did not result in any significant changes in the number of PMN-MDSCs (CD11b+, Ly6G+) (S3 A-C) or mMDSCs (CD11b+, Ly6C+) (S3 D-F) in the bone marrow, spleen or liver. NS = not statistically significant. Error bars represent pooled data from at least 2 independent experiments and are represented as mean ± SEM.
Supplemental Figure S5. PD-L1 expression is not significantly changed on C1498 leukemic blasts following DMXAA + anti-PD-1 therapy. Samples from leukemia engrafted mice were harvested 20 days following treatment with vehicle control, DMXAA, anti-PD-1 antibody or DMXAA + anti-PD-1 antibody. Expression of PD-L1 on C1498 leukemic blasts was measured within (A) the bone marrow, (B) spleen, and (C) liver. NS = not statistically significant. Error bars represent pooled data from at least 2 independent experiments and are represented as mean ± SEM.
Acknowledgements
We thank Dr. Bruce Blazar for kindly providing us with the stably transfected C1498FFDsR cell line, as well as Claire Fritz, Demi Lewis and Dr. Kaitlyn Dykstra for technical assistance. We also thank the Roswell Park Flow Cytometry Core Lab for performing the Luminex assays and for the Roswell Park Pathology Core Lab for performing the immunohistochemistry.
Financial Support:
CCS Grant P30CA016056, Roswell Park Alliance Foundation, Jacquie Hirsch Leukemia Research Fund
Footnotes
Conflicts of Interest
Dr. Wang reports personal fees from Amgen, personal fees from Abbvie, personal fees from Astellas, personal fees from Jazz Pharmaceuticals, personal fees from Pfizer, personal fees from Daiichi Sankyo, personal fees from Arog/Dava Oncology, personal fees from PTC Therapeutics, personal fees from Macrogenics, personal fees from Stemline Therapeutics, personal fees from Kura Oncology, personal fees from Immunogen. The other authors have no relevant conflicts of interest to disclose.
References:
- 1.Ravandi F, Assi R, Daver N, et al. Idarubicin, cytarabine, and nivolumab in patients with newly diagnosed acute myeloid leukaemia or high-risk myelodysplastic syndrome: a single-arm, phase 2 study. Lancet Haematol. 2019. Sep;6(9):e480–e488. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Daver N, Garcia-Manero G, Basu S, et al. Efficacy, Safety, and Biomarkers of Response to Azacitidine and Nivolumab in Relapsed/Refractory Acute Myeloid Leukemia: A Nonrandomized, Open-Label, Phase II Study. Cancer Discov. 2019. Mar;9(3):370–383. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Wang ES. Treating acute myeloid leukemia in older adults. Hematology American Society of Hematology Education Program. 2014. Dec 05;2014(1):14–20. [DOI] [PubMed] [Google Scholar]
- 4.Topalian SL, Drake CG, Pardoll DM. Immune checkpoint blockade: a common denominator approach to cancer therapy. Cancer cell. 2015. Apr 13;27(4):450–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Dolen Y, Esendagli G. Myeloid leukemia cells with a B7–2(+) subpopulation provoke Th-cell responses and become immuno-suppressive through the modulation of B7 ligands. European journal of immunology. 2013. Mar;43(3):747–57. [DOI] [PubMed] [Google Scholar]
- 6.Saudemont A, Quesnel B. In a model of tumor dormancy, long-term persistent leukemic cells have increased B7-H1 and B7.1 expression and resist CTL-mediated lysis. Blood. 2004. Oct 1;104(7):2124–33. [DOI] [PubMed] [Google Scholar]
- 7.Zhang L, Gajewski TF, Kline J. PD-1/PD-L1 interactions inhibit antitumor immune responses in a murine acute myeloid leukemia model. Blood. 2009. Aug 20;114(8):1545–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Zhou Q, Munger ME, Highfill SL, et al. Program death-1 signaling and regulatory T cells collaborate to resist the function of adoptively transferred cytotoxic T lymphocytes in advanced acute myeloid leukemia. Blood. 2010. Oct 7;116(14):2484–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Baguley BC, Ching LM. DMXAA: an antivascular agent with multiple host responses. Int J Radiat Oncol Biol Phys. 2002. Dec 1;54(5):1503–11. [DOI] [PubMed] [Google Scholar]
- 10.Curran E, Chen X, Corrales L, et al. STING Pathway Activation Stimulates Potent Immunity against Acute Myeloid Leukemia. Cell Rep. 2016. Jun 14;15(11):2357–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Zhou Q, Bucher C, Munger ME, et al. Depletion of endogenous tumor-associated regulatory T cells improves the efficacy of adoptive cytotoxic T-cell immunotherapy in murine acute myeloid leukemia. Blood. 2009. Oct 29;114(18):3793–802. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Zhou Q, Munger ME, Veenstra RG, et al. Coexpression of Tim-3 and PD-1 identifies a CD8+ T-cell exhaustion phenotype in mice with disseminated acute myelogenous leukemia. Blood. 2011. Apr 28;117(17):4501–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Baguley BC, Siemann DW. Temporal aspects of the action of ASA404 (vadimezan; DMXAA). Expert Opin Investig Drugs. 2010. Nov;19(11):1413–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Przespolewski AP S; Den Haese J; Wang ES Targeting innate and adaptive immune responses for the treatment of acute myeloid leukemia. [ASH Hematologic Malignancies poster]. 2016. [Google Scholar]
- 15.Sauer MG, Ericson ME, Weigel BJ, et al. A novel system for simultaneous in vivo tracking and biological assessment of leukemia cells and ex vivo generated leukemia-reactive cytotoxic T cells. Cancer research. 2004. Jun 1;64(11):3914–21. [DOI] [PubMed] [Google Scholar]
- 16.Hoebe K, Janssen EM, Kim SO, et al. Upregulation of costimulatory molecules induced by lipopolysaccharide and double-stranded RNA occurs by Trif-dependent and Trif-independent pathways. Nat Immunol. 2003. Dec;4(12):1223–9. [DOI] [PubMed] [Google Scholar]
- 17.Honda K, Sakaguchi S, Nakajima C, et al. Selective contribution of IFN-alpha/beta signaling to the maturation of dendritic cells induced by double-stranded RNA or viral infection. Proceedings of the National Academy of Sciences of the United States of America. 2003. Sep 16;100(19):10872–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Le Bon A, Tough DF. Links between innate and adaptive immunity via type I interferon. Curr Opin Immunol. 2002. Aug;14(4):432–6. [DOI] [PubMed] [Google Scholar]
- 19.Buechler C, Ritter M, Orso E, et al. Regulation of scavenger receptor CD163 expression in human monocytes and macrophages by pro- and antiinflammatory stimuli. J Leukoc Biol. 2000. Jan;67(1):97–103. [PubMed] [Google Scholar]
- 20.Portwood SL D; Schaefer C;, Uwazurike O; Spernyak J; Seshadri M; Wang ES Potent effects of the vascular disrupting agent, ASA404 (DMXAA) on the marrow microenvironment in preclinical human leukemia and lymphoma models. [ASH Poster]. Blood. 2013;122. [Google Scholar]
- 21.Diamond MS, Kinder M, Matsushita H, et al. Type I interferon is selectively required by dendritic cells for immune rejection of tumors. J Exp Med. 2011. Sep 26;208(10):1989–2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Fuertes MB, Kacha AK, Kline J, et al. Host type I IFN signals are required for antitumor CD8+ T cell responses through CD8{alpha}+ dendritic cells. J Exp Med. 2011. Sep 26;208(10):2005–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Gajewski TF. Failure at the effector phase: immune barriers at the level of the melanoma tumor microenvironment. Clinical cancer research : an official journal of the American Association for Cancer Research. 2007. Sep 15;13(18 Pt 1):5256–61. [DOI] [PubMed] [Google Scholar]
- 24.Corrales L, Glickman LH, McWhirter SM, et al. Direct Activation of STING in the Tumor Microenvironment Leads to Potent and Systemic Tumor Regression and Immunity. Cell Rep. 2015. May 19;11(7):1018–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Fu J, Kanne DB, Leong M, et al. STING agonist formulated cancer vaccines can cure established tumors resistant to PD-1 blockade. Sci Transl Med. 2015. Apr 15;7(283):283ra52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Sallets A, Robinson S, Kardosh A, et al. Enhancing immunotherapy of STING agonist for lymphoma in preclinical models. Blood Adv. 2018. Sep 11;2(17):2230–2241. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Ossenkoppele GJ, Stussi G, Maertens J, et al. Addition of bevacizumab to chemotherapy in acute myeloid leukemia at older age: a randomized phase 2 trial of the Dutch-Belgian Cooperative Trial Group for Hemato-Oncology (HOVON) and the Swiss Group for Clinical Cancer Research (SAKK). Blood. 2012. Dec 6;120(24):4706–11. [DOI] [PubMed] [Google Scholar]
- 28.Herbst RS, Soria JC, Kowanetz M, et al. Predictive correlates of response to the anti-PD-L1 antibody MPDL3280A in cancer patients. Nature. 2014. Nov 27;515(7528):563–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Daei Farshchi Adli A, Jahanban-Esfahlan R, Seidi K, et al. An overview on Vadimezan (DMXAA): The vascular disrupting agent. Chem Biol Drug Des. 2018. May;91(5):996–1006. [DOI] [PubMed] [Google Scholar]
- 30.Cerneski SAM, Bandi M, Chang W, Chen Y, Long C, et al. Preclinical characterization of a novel STING agonist, MK-1454 [Abstract ]. 2017. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplemental Figure S1. DMXAA results in vascular disruption within the bone marrow. Samples were harvested from leukemia engrafted animals 20 days following treatment with vehicle control or single agent DMXAA. Bone marrow sections were stained with VEGFR. Representative photomicrographs demonstrate increased vascular disruption as measured by increase in vessel diameter. Error bars represent pooled data from at least 2 independent experiments and are represented as mean ± SEM.
Supplemental Figure S2. DMXAA + anti-PD-1 therapy does not impact regulatory T-cell populations. Samples were harvested from leukemia engrafted animals 20 days following treatment with vehicle control, DMXAA, anti-PD-1 antibody or DMXAA + anti-PD-1 antibody. A trend towards decreased regulatory T-cells (CD4+, CD25+, FOXP3+) was not observed in the (A) bone marrow but was observed in the (B) spleen and (C) liver of leukemic mice. NS = not statistically significant. Error bars represent pooled data from at least 2 independent experiments and are represented as mean ± SEM.
Supplemental Figure S3. DMXAA + anti-PD-1 therapy recruits more CD163+ macrophages. Samples were harvested from leukemia engrafted animals 20 days following treatment with vehicle control, DMXAA, anti-PD-1 antibody or DMXAA + anti-PD-1 antibody. Representative photomicrographs demonstrate CD163+ macrophages were found to be present at increased frequency in the spleen of animals treated with combination therapy (D), compared to single agent anti-PD-1 ab (C), DMXAA (B) or vehicle control (A).
Supplemental Figure S4. DMXAA + anti-PD-1 therapy does not affect populations of myeloid-derived suppressor cells. Samples were harvested from leukemia engrafted animals 20 days following treatment with vehicle control, DMXAA, anti-PD-1 antibody or DMXAA + anti-PD-1 antibody. Treatment did not result in any significant changes in the number of PMN-MDSCs (CD11b+, Ly6G+) (S3 A-C) or mMDSCs (CD11b+, Ly6C+) (S3 D-F) in the bone marrow, spleen or liver. NS = not statistically significant. Error bars represent pooled data from at least 2 independent experiments and are represented as mean ± SEM.
Supplemental Figure S5. PD-L1 expression is not significantly changed on C1498 leukemic blasts following DMXAA + anti-PD-1 therapy. Samples from leukemia engrafted mice were harvested 20 days following treatment with vehicle control, DMXAA, anti-PD-1 antibody or DMXAA + anti-PD-1 antibody. Expression of PD-L1 on C1498 leukemic blasts was measured within (A) the bone marrow, (B) spleen, and (C) liver. NS = not statistically significant. Error bars represent pooled data from at least 2 independent experiments and are represented as mean ± SEM.