A Combined PD-1/C5a Blockade Synergistically Protects against Lung Cancer Growth and Metastasis

Immunotherapy based on checkpoint inhibitors has emerged as a potent tool for the treatment of lung cancer, the leading cause of cancer-related death worldwide. Among the numerous immunotherapeutic strategies, monoclonal antibodies that inhibit the interaction between programmed death-1 (PD-1) and its ligands have shown the most compelling clinical results in lung cancer. Monoclonal antibodies raised against PD-1 have been approved by the FDA for patients with metastatic non–small cell lung cancer (1). By blocking PD-1, these drugs remove the inhibition of T-cell activation and restore antitumor immune responses. However, PD-1 inhibition is not capable of reversing all resistance mechanisms, and a proportion of patients do not respond adequately to anti–PD-1 immunotherapies. Consequently, combination therapies, which block more than one immunomodulatory pathway, have been proposed to further enhance the antitumor efficacy of anti–PD-1 individual treatments (2).

A hallmark mechanism of synergy in immunotherapy is the reactivation of effector T cells together with the elimination of immunosuppressive cells, such as myeloid-derived suppressor cells (MDSC) or regulatory T cells (Treg). In this regard, PD-1/PD-L1 blockade is not able to reduce the T-cell suppressive activity of the tumor microenvironment, which can be caused by the accumulation of MDSCs (3) or Tregs (4). Therefore, a successful reversal of the tumor immunosuppressive microenvironment represents a window of opportunity to overcome tumor resistance in cancer immunotherapy.

Anaphylatoxin C5a, an active proteolytic fragment released after activation of the complement system, contributes to lung cancer progression by promoting an immunosuppressive microenvironment in which MDSCs are involved (5, 6). All pathways of complement, an essential part of innate immunity, converge in the cleavage and activation of the central components C3 and C5, which leads to the release of the anaphylatoxins C3a and C5a. Although C5a is generally known to be a chemoattractant for proinflammatory leukocytes, several studies have demonstrated a tumor-promoting role for this molecule, and the idea of blocking complement for the treatment of cancer is gaining recognition (7). In mouse models of cancer, complement deficiencies and pharmacologic blockade of complement-related mediators, including C5a, have been associated with impaired tumor growth (8). C5a is also able to modulate antitumor immunity by altering T-cell responses in premetastatic niches (9). In the case of lung cancer, the expression of some immunosuppressive molecules is significantly reduced after blockade of the C5a receptor-1 (C5aR1; ref. 5). Besides, C5aR1 blockade diminishes the percentage and activity of MDSCs in cancer models, including lung cancer (5, 8, 10). By studying C3 knockout mice, it has been recently shown that complement inhibits antitumor immunity through a PD-1 independent pathway (11). All of these studies reveal the important role played by C5a/C5aR1 signaling in tumor immunity and point to this pathway as a potential therapeutic target in the context of checkpoint inhibition.

In this study, we have evaluated the therapeutic efficacy of the combined administration of anti–PD-1 and anti-C5a drugs in a variety of lung cancer models. From these experiments, we conclude that C5a blockade results in a more permissive environment for immune-mediated tumor rejection and, consequently, its combined administration with anti–PD-1 antibodies synergistically impairs lung cancer growth and metastasis.

To explore the role of C5a/C5aR1 signaling in lung cancer initiation and progression, we crossed C5aR1-deficient mice (C5aR1^+/−) to Kras^LSL-G12D/+ mice that develop lung adenocarcinomas by inducing the expression of oncogenic Kras upon intratracheal inoculation of Cre recombinase. Deletion of C5aR1 had a noticeable effect on mutant Kras–induced lung carcinogenesis (Fig. 1A). Thus, Kras^LSL-G12D/+;C5aR1^−/− mice developed significantly smaller tumors compared with C5aR1 wild-type littermates, whereas no differences were observed in the total number of tumors (Fig. 1B). Kras^LSL-G12D/+;C5aR1^−/− mice showed a reduced percentage of MDSCs in their spleens (Fig. 1C). No differences were observed in the frequencies of CD4 T cells, CD8 T cells, or Tregs (data not shown). Immunohistochemical analysis of the tumors from Kras^LSL-G12D/+; C5aR1^−/− mice showed an increase in CD3 cells close to statistical significance (Fig. 1D). No differences were found in proliferation (Ki67), apoptosis (cleaved caspase-3), or angiogenesis (CD31; data not shown). Using an independent mouse cohort, we observed that survival of Kras^LSL-G12D/+;C5aR1^−/− mice was significantly longer than that of Kras^LSL-G12D/+;C5aR1^+/+ littermates (Fig. 1E).

We next investigated the effect of the pharmacologic blockade of C5a in a syngeneic mouse model based on the subcutaneous growth of 393P cells, a mouse model of KRAS-driven lung adenocarcinoma (12). To block C5a, we used AON-D21 (formerly NOX-D21), an l-aptamer that tightly binds to both mouse and human C5a and efficiently inhibits the interaction with its receptors (Supplementary Results, Supplementary Fig. S1, and Supplementary Table S1 show information about the pharmacology and other characteristics of this molecule). AON-D21 treatment led to a partial reduction of tumor growth when compared with the effect of revAON-D21, a nonfunctional control l-aptamer (Fig. 1F). Analysis of splenocyte subpopulations revealed a reduction in the percentage of MDSCs (Fig. 1G), with no significant differences in CD4 T cells, CD8 T cells, or Tregs (data not shown). These genetic and pharmacologic results demonstrate the influence of the C5a/C5aR1 pathway on lung tumor growth and the potential value of blocking C5a for the treatment of lung cancer.

We postulated that an anti-C5a treatment may enhance the capacity of PD-1 blockade to control lung tumor growth. In the subcutaneous 393P model, the combination of the anti-C5a l-aptamer AON-D21 (every other day) with the anti–PD-1 monoclonal antibody RMP1-14 (days 7, 10, and 14) resulted in a significant reduction of tumor growth, as compared with the effect of each treatment alone (Fig. 2A and Supplementary Fig. S2A). By day 22, tumor volumes in mice treated with the combined therapy were lower than those in control mice, mice treated with the anti–PD-1 antibody or treated with the anti-C5a aptamer. Remarkably, by day 41, all the mice (n = 9) in the combination group showed complete tumor rejection. In contrast, at day 48, only one mouse in the control group, two in the group treated with AON-D21, and four in the group treated with anti–PD-1 showed tumor rejection. Moreover, all these mice resisted a tumor rechallenge performed after an untreated period of 80 days, suggesting the development of an efficient antitumor long-term memory response (Supplementary Fig. S2B). We also tested the effect of the combination in a different syngeneic mouse model based on the subcutaneous implantation of Lewis lung adenocarcinoma (LLC) cells (Fig. 2B and Supplementary Fig. S2C). Anti–PD-1 treatment on days 7, 10, and 14 administered in combination with AON-D21 (every other day) resulted in an attenuation of tumor growth that became significant by day 14. Although complete rejections were not observed in this model, a survival analysis confirmed the benefit of the combined treatment (Fig. 2C). The combination treatment was also able to increase the survival of mice following injection of LLC cells into the tail vein (Fig. 2D), suggesting that the treatment inhibited lung cancer metastasis. Lastly, C5a blockade synergized with PD-1 blockade to prevent multiorgan metastases after intracardiac injection of highly aggressive Kras-mutated lung adenocarcinoma Lacun3 cells (Fig. 2E). A survival analysis performed in an independent experiment in which anti–PD-1 treatment was ceased on day 10 confirmed the benefit of the combined treatment (Fig. 2F). In conclusion, the combined immunotherapy based on C5a and PD-1 blockade showed synergistic effects on both lung cancer growth and metastatic progression.

To identify changes induced in the tumor microenvironment by C5a and/or PD-1 blockade, we analyzed the immune cells present in 393P tumor–bearing mice. By day 14, statistically significant differences in tumor size were observed between groups and were maintained until the end of the experiment (Fig. 3A). Animals were sacrificed at day 26 before tumors were completely rejected by the combined treatment. At this point, the combined anti-C5a/PD-1 treatment resulted in a significant elevation of the percentage of CD8 T cells in the spleens of the treated animals, as compared with the control mice, whereas no changes were observed in other immune subpopulations (Supplementary Fig. S3). In tumors, the effect on the frequency of CD8 T cells was stronger (Fig. 3B and Supplementary Fig. S4). In addition, the combination treatment led to a decrease in the frequency of MDSC leukocyte subpopulation in tumors, as compared with the control group (Fig. 3B). None of the treatments significantly modified the frequency of CD4 T cells, natural killer (NK) cells, or Tregs in tumors, except for a reduction of NK cells in tumors treated with the dual blockade (Supplementary Fig. S5). Further analyses revealed that in tumors treated with the dual therapy, there was a negative correlation between CD8 T cells and MDSCs, as well as an increase in CD45-positive cells (Fig. 3B). In support of the relevance of MDSCs in the model, anti–PD-1 treatment in combination with MDSC depletion had a therapeutic effect similar to that observed with the anti–PD-1/anti-C5a treatment (Fig. 3C). In relation to the involvement of potential effector cells, a selective depletion of CD8 T cells completely abrogated the antitumor efficacy of the combination therapy against 393P cells (Fig. 3D). CD4 T cells and NK cells seemed to be dispensable, although in the case of NK cells a moderate growth increase was observed at the last days of the experiment, suggesting that NK cell depletion may partially hamper the antitumor effect of the therapy. Tumor-infiltrating immune cells were also analyzed in mice bearing LLC tumors treated with anti-C5a and/or anti–PD-1 agents. Results were similar to those previously obtained with 393P tumors, although in this case significant differences were not always found (Supplementary Figs. S6, S7A, and S7B).

We also analyzed the mRNA expression in the tumors of a battery of immune-related mediators. In 393P tumors, we found significant differences in the expression of three molecules: IL2, LAG3, and CCL17 (Fig. 3E and Supplementary Table S2). In the LLC model, mRNA expression of the T-cell activating cytokine IL2 was consistently elevated after the combination treatment, albeit without reaching statistical significance (Supplementary Fig. S7C). Results for LAG3 and CCL17 were not consistent with those obtained in the 393P model (data not shown). From all of these experiments, we conclude that the anti–PD-1/anti-C5a combination synergistically impairs tumor growth by the action of CD8 T cells in association with an elevation of IL2.

To better evaluate the potential of combined therapy in a model that more closely resembles the clinical setting, we treated established 393P tumors during a defined period of time. We observed regressions in 7 of 8 mice treated with the combination of anti-C5a (days 9, 10, and every other day until day 24) and anti–PD-1 (days 11, 14, and 18) drugs (Fig. 4A). The combined treatment resulted in a significant improvement in survival, despite the fact that after treatment termination most tumors started growing again (Fig. 4B). In an independent experiment, we analyzed the expression of surface activation markers in tumor-infiltrating CD8 T cells. In agreement with previous results, we observed an increase in the percentage of tumor CD8 T cells (Fig. 4C). In addition, these cells showed a marked reduction in the exhaustion markers PD-1, GITR, and LAG3 (Fig. 4C). PD-1 was already downregulated in those tumors treated with anti–PD-1 alone; however, the expression of the two other markers was reduced only upon the administration of the combination therapy. We also found a trend of an increase in IL2 protein expression within the tumors (Fig. 4D). In conclusion, the combination of anti-C5a/anti–PD-1 drugs shows a significant therapeutic efficacy on tumor-bearing mice. This activity seems to be associated with an increase in the expression of IL2 and an attenuated CD8 T-cell dysfunction as compared with the anti–PD-1 monotherapy.

Numerous studies are under way to identify synergistic combinations of checkpoint inhibitors with chemotherapy, radiotherapy, targeted therapy, or other immunotherapy strategies (2). Many of these combinations are built on PD-1/PD-L1 inhibition, for which the presence of immunosuppressive pathways within the tumor microenvironment represents a major hurdle. In this study, we provide the framework for the clinical evaluation of an innovative combination strategy in which anti–PD-1 antibodies are administered together with drugs that inhibit C5a signaling pathways. Overall, our study demonstrates that C5a blockade synergizes with anti–PD-1 inhibition in preclinical models of local and metastatic lung cancer growth. The rationale behind this approach is that the activation and expansion of effector T cells can be optimized by targeting immunosuppressive cells in the tumor microenvironment.

Previous studies performed in diverse mouse models have led to the proposal that C5a/C5aR1 inhibition may represent a novel therapeutic target for cancer (13). Among the different potential mechanisms, C5a/C5aR1 blockade could partially reverse the immunosuppressive tumor microenvironment linked to the activity of MDSCs (5, 8, 10). Nonetheless, in the preclinical lung cancer models of our study the anti-C5a l-aptamer AON-D21 used as monotherapy showed only modest antitumor effects. In contrast, its administration in combination with an anti–PD-1 antibody greatly influenced the immune response in a way that resulted in remarkable antitumor activity, at both primary lung tumors and metastatic sites.

The most plausible explanation for the synergism between anti-C5a and anti–PD-1 drugs is that C5aR1 signaling mediates mechanisms that hamper the antitumor activity of anti–PD-1 antibodies. In agreement with this postulate, our data show that the combined treatment leads to a lower frequency of MDSCs and a higher frequency of CD8 T cells associated with a decreased expression of exhaustion markers, which suggests a more complete restoration of CD8 T-cell effector functions. The combination treatment was also associated with an increase in the levels of IL2, a pleiotropic cytokine with a critical role in multiple aspects of CD8 T-cell activation (14). The essential role played by CD8 T cells was reinforced by the complete abrogation of the immunotherapeutic effect after depletion of these cells. Previous studies had already suggested a role of complement in the activity of tumor-infiltrating CD8 T cells. Induction of a tumor-specific CD8 T-cell response was improved upon transient inhibition of the complement system (15), and a simultaneous blockade of C3aR and C5aR1 inhibited the development of breast tumors by enhancing the effector capacity of CD8 T cells (11). Interestingly, in the same study, a combined inhibition of C3aR/C5aR1 and PD-1 enhanced antitumor effects (11). Considering these studies, a direct effect of C5a blockade on CD8 T cells is also a plausible mechanism in our models, which may lead to CD8 T-cell expansion. Contrastingly, in other pathologic conditions, C5aR1 signaling seemed to be essential for an optimal generation of CD8 T-cell responses (16). In regard to the potential role played by MDSCs, their contribution to the antitumor activity of the combined therapy is supported by the reduction in the percentage of these cells in some of the studied models, its negative correlation with CD8 T cells, and the biological relevance of these cells in these models. Nevertheless, we cannot completely exclude the possibility that the reduced frequency of MDSCs was a bystander finding in our experiments, because C5a inhibition, alone or in combination with PD-1 inhibition, did not always lead to significant changes. In-depth mechanistic studies are still required to properly assess the roles played by MDSCs and CD8 T cells, without excluding the involvement of other immune cells, such as NK cells or monocytes. Interestingly, a recent study has demonstrated that C5a promotes immunosuppressive responses by contributing to the expression of PD-L1 on monocytes (17).

A comparison of the two subcutaneous tumor growth models used in our study suggests a more effective antitumor immune response in the KRAS-driven model of lung adenocarcinoma. Consequently, the impact of Kras mutations on the antitumor efficacy of the combination should be addressed, as well as the potential influence of other genetic determinants leading to neoantigen generation. Further mechanistic studies would also provide guidance for the most efficient administration schedule for immune stimulation, as well as for the potential benefit of blocking other complement mediators, such as C3aR (11). The impact of the alternative receptor for C5a, C5aR2 (C5L2), which functions are still poorly understood, should also be investigated. Finally, the extension of this combination to other tumor types merits further evaluation. Its applicability to models of lung squamous cell carcinoma, the second most common subtype of lung cancer for which anti–PD-1 therapies have also shown clinical utility (1), deserves particular consideration.

A major concern in the application of immunotherapy combinations is the potentiation of adverse effects as a result of an excessive immune activation (2). It is premature to speculate about the safety of the combination presented in this study, but clinical data from the use of the C5-blocking antibody eculizumab show that safety issues associated with complement inhibition are mainly related to the blockade of C5b-mediated bacterial lysis (18). AON-D21 selectively blocks C5a and does not interfere with C5b biology despite binding to intact C5 (19, 20). Such a selective blockade of C5a/C5aR1 was safe and generally well tolerated in clinical trials (21, 22). The substance class of l-aptamers (Spiegelmers) was also generally safe and well tolerated in clinical phase I and IIa studies (23).

In conclusion, our work supports the notion that the efficacy of anti–PD-1 therapies is limited by their inability to fully reprogram the tumor microenvironment, which maintains other immunosuppressive mediators. In this context, C5a/C5aR1 blockade overcomes some of the resistance mechanisms, markedly improving antitumor immune responses. These findings provide support for the clinical evaluation of anti–PD-1 and anti-C5a drugs as a combination therapy for lung cancer.

The 393P cells, derived from Kras^LA1/+;p53^R172HΔG mice, were a generous gift from Dr. J.M. Kurie (The University of Texas MD Anderson Cancer Center, Houston, TX). LLC cells were obtained from the American Type Culture Collection. The Lacun3 cell line was previously established by our group from a chemically induced lung adenocarcinoma and stably transfected with the luciferase gene (24). Nontransduced Lacun3 cells were used for the survival experiment. Cells were grown in RPMI-1640 supplemented with 2 mmol/L glutamine, 10% Fetalclone (Thermo), 100 U/mL penicillin, and 100 mg/mL streptomycin (Invitrogen). All cell lines were routinely tested for Mycoplasma. Cell line authentication was not routinely performed.

AON-D21, a follow-up of previously described NOX-D19 and NOX-D20 (19, 20), is a novel PEG-modified l-aptamer (Spiegelmer). revAON-D21, a Spiegelmer of the reverse AON-D21 sequence, was used as a negative control. Both aptamers were synthesized at NOXXON Pharma. Methods for the evaluation of AON-D21 affinity, pharmacokinetics, and inhibitory capacity are described in the Supplementary Methods section.

All animal experiments were conducted in accordance with the protocols approved by the Institutional Animal Care Committee of the University of Navarra. Kras^LSL-G12D/+;C5aR1^+/− animals were intercrossed to generate Kras^LSL-G12D/+;C5aR1^+/+ or Kras^LSL-G12D/+;C5aR1^−/− offspring. Kras^LSL-G12D/+ mice harbor a conditionally activatable allele of oncogenic mutant Kras. Intratracheal inoculation with adenoviral Cre (AdCre) led to expression of mutant Kras. Animals were followed for a period of at least 4 months. Murine 393P cells or LLC cells (1.5 × 10⁶) were resuspended in 50 μL of PBS and injected subcutaneously in the right flanks of 8-week-old Sv/129 or C57BL/6J mice, respectively. One day before and on the day of cell injection, mice were treated with AON-D21 or revAON-D21 (10 mg/kg in saline, i.p.), and treatment was continued every other day until the end of the experiment. For combination therapy, tumor-bearing mice were treated with anti–PD-1 blocking antibody (RMP1-14, BioXCell) at days 7, 10, and 14 after cell inoculation (100 μg per mouse in PBS, i.p.). Depletion of CD8⁺, CD4⁺, or NK⁺ cells was achieved by intraperitoneal injection of 100 μg of antimouse CD8α (clone 2.43; BioXCell), CD4 (clone GK1.5; BioXCell), or NK1.1 (clone PK136; BioXCell), respectively, at days 6, 11, 14, 18, 21, and 28 after cell inoculation. Irrelevant IgG (BioXCell) was administered as control. MDSCs were depleted by intraperitoneal injection of 200 μg of antimouse anti–Gr-1 (Ly6G/C; clone RB6-8C5; BioXCell) on the day of cell injection and every other day until the end of the experiment. Tumors were measured periodically and volumes were calculated by the formula (L × W²)/2, where L is the length and W is the width. Animals were euthanized when tumor diameters reached 17 mm or when they appeared under distress. The metastasis model based on LLC cells was performed by intravenous injection of 1 × 10⁶ cells into the tail vein of C57BL/6J mice. The multiorgan metastatic potential of luciferase-transfected Lacun3 cells was evaluated after inoculation of 2 × 10⁵ cells into the left cardiac ventricle of BALB/c mice as previously described (24, 25). In these two metastasis models, treatments were performed as described above, except for the injections of anti–PD-1 blocking antibody, which were performed at days 3, 7, and 10 after cell inoculation. Finally, the experiments performed with established tumors (Fig. 4) were carried out as follows: subcutaneously inoculated 393P cells (1.5 × 10⁶) were allowed to grow for 9 days, mice were randomized, and treatment started with AON-D21 (10 mg/kg, i.p.) at days 9 and 10 and every other day until day 24. Anti–PD-1 blocking antibody (100 μg per mouse) was administered at days 11, 14, and 18 after cell inoculation.

Tumors and spleens from tumor-bearing mice were mechanically disaggregated as previously described (5). Erythrocytes were lysed in a buffer containing 155 mmol/L NH₄Cl and 10 mmol/L KHCO₃. Single-cell suspensions were treated with Fc block (2.4G2; BD Pharmingen) and then stained with a labeled primary antibody against mouse CD45 (30-F11; BioLegend), CD8a (53-6.7; BD Pharmingen), NK1.1 (PK136; BioLegend), CD11b (M1/70; BioLegend), Ly6C (AL21; BD Pharmingen), Ly6G (1A8; BioLegend), LAG3 (C9B7W; BioLegend), PD-1 (29F.1A12; BioLegend), and GITR (YGITR 765; BioLegend) diluted in FACS buffer (PBS, 0.1% NaN₃, 1% BSA). Staining of CD4 T cells and Tregs was performed using a kit from eBioscience according to the manufacturer's instructions. As an example, the gating strategy for MDSCs and CD8 T cells is shown in Supplementary Fig. S8. When indicated, dead cells were excluded using the Zombie NIR Fixable Viability Kit (BioLegend), and the absolute number of CD8 T cells per milligram of tumor was determined using Cytognos Perfect Microspheres (Cytognos). Cells were acquired using a BD Biosciences FACSCalibur flow cytometer, except for the analysis of LAG3, GITR, and PD-1 expression on CD8 T cells, which was performed on a FACSCANTO II flow cytometer. Data were analyzed using BD CellQuest Pro (BD Biosciences) and FlowJo software (TreeStar).

Immunohistochemistry was performed on formalin-fixed paraffin-embedded tissue sections. After antigen retrieval with EDTA buffer (CD3 and Ki67) or citrate buffer (cleaved caspase-3 and CD31), sections were incubated with the primary antibodies (anti-CD3, 1:300, Thermo Scientific; anti-Ki67, 1:100, Neomarkers; anti-cleaved caspase-3, 1:100, Cell Signaling; anti-CD31, 1:20, Dianova) overnight at 4°C, and revealed with the EnVision HRP System (Dako) and diaminobenzidine. Quantification of staining was performed either automatically (Ki67, cleaved caspase-3, and CD31) or manually (CD3).

Portions of ∼0.1 cm³ were cut from the edge of tumors, frozen in dry ice, and maintained at −80°C until extraction. mRNA expression was evaluated by real-time PCR as previously described (5). For protein analysis, frozen tumors were lysed in an extraction buffer (100 mmol/L Tris, 150 mmol/L NaCl, 1 mmol/L EGTA, 1 mmol/L EDTA, 1% Triton X-100, 0.5% sodium deoxycholate, pH 7.4) containing a cocktail of protease inhibitors. Extracts were analyzed using the OptEIA mouse IL2 ELISA kit (BD Biosciences).

For dot plots, individual results and median per group are shown. For tumor volumes, the mean ± SEM is depicted. Comparisons between two groups were performed using the Mann–Whitney U test. Comparisons between treatment strategies and between immune-regulator levels were performed using the Kruskal–Wallis test with the Mann–Whitney U test as the post hoc test. Survival curves were generated using the Kaplan–Meier method, and differences were analyzed with the log rank test. For these analyses, survival times were defined as the number of days from the inoculation of the cells until the mice were euthanized or expired naturally. The Spearman rank correlation test was performed to analyze association. Significances in tumor volumes and survival tests were always calculated versus the control group. P < 0.05 was considered statistically significant. Statistical analyses were performed using Prism software (GraphPad).

K. Hoehlig and A. Vater are authors of patents and patent applications claiming certain anti-C5a L-aptamers, including AON-D21. A. Vater has ownership interest in Aptarion Biotech AG and NOXXON Pharma AG. No potential conflicts of interest were disclosed by the other authors.

Conception and design: D. Ajona, A. Vater, L.M. Montuenga, R. Pio

Development of methodology: D. Ajona, H. Moreno, M.J. Pajares, J. Agorreta, S. Vicent, F. Lecanda, R. Pio

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): D. Ajona, S. Ortiz-Espinosa, H. Moreno, T. Lozano, M.J. Pajares, J. Agorreta, C. Bértolo, J.J. Lasarte, S. Vicent, K. Hoehlig

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): D. Ajona, S. Ortiz-Espinosa, H. Moreno, T. Lozano, J. Agorreta, C. Bértolo, J.J. Lasarte, K. Hoehlig, F. Lecanda, L.M. Montuenga, R. Pio

Writing, review, and/or revision of the manuscript: D. Ajona, S. Ortiz-Espinosa, J. Agorreta, C. Bértolo, J.J. Lasarte, S. Vicent, K. Hoehlig, A. Vater, F. Lecanda, L.M. Montuenga, R. Pio

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): D. Ajona, J. Agorreta, C. Bértolo, A. Vater, R. Pio

Study supervision: D. Ajona, J. Agorreta, R. Pio

We are grateful to Ignacio Melero for critically reviewing the manuscript. We thank Carolina Zandueta, Cristina Sainz, and Amaya Lavín for technical assistance; Jonathan M. Kurie (The University of Texas MD Anderson Cancer Center, Houston, TX) for the generous gift of 393P cells; and Diego Alignani (Cytometry Platform, CIMA), Noelia Casares, and Sandra Hervas-Stubbs (Program in Immunology and Immunotherapy, CIMA) for their help with flow cytometry. The authors would also like to thank the following individuals at NOXXONPharma: Lucas Bethge (chemical synthesis of AON-D21, revAON-D21, and hybridization probe for bioanalysis), Christian Maasch (AON-D21 affinity determination and bioanalysis from mouse pharmacokinetics by surface plasmon resonance experiments), and Klaus Buchner and Dirk Zboralski (cell-based potency assays).

This work was supported by the Foundation for Applied Medical Research (FIMA), Ramon Areces Foundation, Red Temática de Investigación Cooperativa en Cáncer (RD12/0036/0040), CIBERONC (CB16/12/0043), Fondo de Investigación Sanitaria-Fondo Europeo de Desarrollo Regional (FEDER; PI13/00806 and PI14/01686), the Spanish Ministry of Economy and Competitiveness (SAF2013-46423-R, SAF2013-42772-R, SAF2016-78568-R, and SAF2015-71606-R), and the European Commission (618312 KRASmiR FP7-PEOPLE-2013-CIG). S. Vicent is a fellow of the Ramón y Cajal Program (MICINN, RYC-2011-09042). F. Lecanda is funded by “la Caixa” Foundation and Caja Navarra Foundation. S. Ortiz-Espinosa is a fellow of the Asociación Amigos de la Universidad de Navarra.

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.