The Angiotensin Receptor Blocker Losartan Suppresses Growth of Pulmonary Metastases via AT1R-Independent Inhibition of CCR2 Signaling and Monocyte Recruitment Free

Metastasis remains the greatest clinical challenge in cancer treatment, accounting for up to 90% of all cancer-related deaths (1, 2). For example, for breast and colorectal cancer, distant metastases are present in 6–21% of patients at the time of diagnosis, respectively (3). Furthermore, although 5-y survival rates experienced by breast and colorectal cancer patients with localized disease are typically excellent, these individuals still have a substantially increased lifetime risk for metastasis, with 30–50% eventually developing disseminated disease (4–6). Thus, the development of new therapies that halt metastatic progression remains a critical hurdle in improving patient outcome.

The tumor microenvironment (TME) is comprised of highly heterogeneous populations of both stromal and immune cells whose diverse functions collectively promote tumor growth, progression, and eventual metastasis (7, 8). Inflammatory monocytes (IMs) are one component of the TME and have recently been shown to play key roles in the metastatic process (9). For example, CCR2-expressing IMs have been shown to be preferentially recruited early to metastatic sites, such as the lung and liver, via tumor cell– and stromal cell–mediated production of the monocyte chemoattractant cytokine, CCL2 (10, 11). At sites of metastases, IMs and their derivatives, metastasis-associated macrophages (MAMs), play key roles in promoting metastatic tumor cell extravasation and growth (10, 12–14). In addition, multiple clinical studies have demonstrated a negative prognostic role for increased numbers of IMs and elevated serum CCL2 concentrations in patients with various malignancies, including those of the breast, colon, and pancreas (15–18). Thus, IMs and the CCL2–CCR2 chemotactic axis represent an attractive target for the treatment of cancer metastasis.

Initial clinical trials targeting the CCL2–CCR2 axis in human cancer patients evaluated an anti-human CCL2 mAb (carlumab, CNTO888) and showed that CNTO888 alone or in combination with standard of care therapies was ineffective at slowing tumor progression in patients with various solid tumors (19). However, more recent trials have shown that the blockade of the CCL2 receptor, CCR2, has the potential to suppress tumor growth in patients with bone metastases and locally advanced pancreatic cancer, suggesting that the inhibition of CCR2 might be a more effective approach in the therapeutic targeting of the CCL2–CCR2 axis (20). Although these recent early phase I/II trials of CCR2 inhibitors are promising, approval is far from certain as recent data suggest that only 10% of agents entering clinical cancer trials make it to Food and Drug Administration approval (21, 22). Furthermore, the time and cost invested in new drugs is now estimated at a staggering 10 y and $2.6 billion (23). Thus, alternative drug development programs that focus on repurposing already approved drugs as potential anticancer therapies offer greater promise, in terms of reduced cost and time, for getting more effective treatment options to cancer patients (24).

Losartan, a type I angiotensin II receptor (AT1R) blocker used in the treatment of hypertension, has been shown to have immunomodulatory and anti-inflammatory properties in models of vascular inflammation and multiple sclerosis (25–27). Interestingly, these anti-inflammatory properties were primarily associated with reduced monocyte and macrophage recruitment to inflammatory lesions and atherosclerotic plaques (25–27). In those studies, losartan blockade of monocyte and macrophage recruitment was attributed to the primary inhibition of angiotensin II–AT1R signaling. However, the impact of losartan on CCL2–CCR2 signaling was not investigated. In fact, some molecular modeling studies suggest that losartan and other angiotensin receptor blockers (ARBs) have the potential to act as direct CCR2 antagonists (28). Losartan has been investigated for the treatment of orthotopic tumors in mice but has not to date been investigated for treatment of tumor metastasis (29, 30). Moreover, the molecular pharmacology underlying losartan’s interactions with CCR2, and its potential to act as a CCR2 antagonist, has not been evaluated.

Therefore, in the current work, we evaluated the ability of losartan to directly inhibit monocyte migration and recruitment using a combination of in vitro and in vivo assays of monocyte chemotaxis, monocyte responses to acute inflammation, and monocyte recruitment in early tumor metastasis. Our results demonstrated that both losartan and its primary metabolite (EXP-3174) potently inhibited CCL2–CCR2 dependent recruitment of human and murine monocytes at clinically relevant concentrations. Furthermore, using G-protein coupled receptor function assays, we characterized the effects of losartan and EXP-3174 on CCL2 ligand binding and postreceptor signal transduction pathways stimulated by CCL2.

Our findings indicated that losartan inhibited CCR2 signaling in a noncompetitive manner in a process independent of the effects on AT1R signaling. In experimental metastasis models of breast and colon cancer, losartan treatment significantly slowed metastatic progression, an effect associated with the blockade of IM recruitment and the reduction in MAMs and tumor angiogenesis. Taken together, these studies indicate that losartan (and potentially other ARB drugs) represent a novel class of safe and approved, noncompetitive negative modulators of CCR2 signaling, which could be efficiently repurposed for use in combination therapy for the prevention or treatment of metastatic disease.

Six- to eight-week-old female BALB/c and ICR mice were purchased from Harlan Laboratories (Denver, CO). CCR2^−/− and Agtr1^−/− (AT1R^−/−) mice on the C57BL/6J background and wild-type C57BL/6J mice were purchased from The Jackson Laboratory (Bar Harbor, ME). CCR2^−/− mice on a BALB/c background were obtained from Dr. C. Ju (University of Colorado, Denver, CO).

4T1-luc, CT26, and CT26-luc cells were generously provided by Dr. D. Gustafson (Colorado State University, Fort Collins, CO). THP-1 cells were purchased from the American Type Culture Collection. CT26-GFP cells were generated by transducing CT26 cells with lentivirus particles expressing GFP under the EF1A promoter (LVP425; GenTarget, San Diego, CA). After 72 h, cells were treated with G-418 (600 μg/ml; InvivoGen, San Diego, CA) to select for successfully transduced cells, and GFP expression was subsequently confirmed by flow cytometry. Cells were maintained in MEM (4T1 and CT26) or RPMI 1640 (THP1) media (Life Technologies, Grand Island, NY) supplemented with 10% FBS (Atlas Biologicals, Fort Collins, CO), penicillin (100 U/ml), streptomycin (100 μg/ml), l-glutamine (2 mM), and nonessential amino acids (0.1 mM) (all obtained from Life Technologies). Cells were grown sterilely on standard plastic tissue culture flasks (CELLTREAT Scientific Products, Shirley, MA) under standard conditions of 37°C, 5% CO₂, and humidified air and were confirmed Mycoplasma-free. CCR2 expression by THP-1 cells was periodically confirmed by flow cytometry.

Fifty milligrams losartan potassium tablets (Cozaar) were obtained from the Veterinary Teaching Hospital pharmacy, ground using a mortar and pestle, and dissolved in water and sterile filtered to obtain a stock concentration of 10 mg/ml. Losartan carboxylic acid (EXP-3174 metabolite) was purchased as a powder from Santa Cruz Biotechnology (Dallas, TX) and reconstituted in DMSO at 10 mg/ml. INCB3284 and RS102895 powder stocks were obtained from Tocris Bioscience (Bristol, U.K.) and reconstituted in DMSO at 10 mg/ml. For all animal experiments, losartan and losartan EXP-3174 metabolite drug stocks were diluted in PBS and administered by once daily i.p. injection of 60 and 10 mg/kg, respectively, in a 100-μl vol.

Wild-type or CCR2^−/− BALB/c mice were inoculated by i.v. tail vein injection of 1 × 10⁵ 4T1-luc cells, 2.5 × 10⁵ CT26-luc cells, or 4 × 10⁵ CT26-GFP cells in 100 μl PBS. Treatment with losartan (60 mg/kg, i.p.) was initiated 24 h after tumor cell inoculation. For 72 h metastasis assays, mice were treated a total of three times (24, 48, and 72 h) prior to euthanasia and tissue collection. For long-term tumor growth studies, mice were treated daily until study completion. To monitor the development and growth of luciferase-positive pulmonary metastases, bioluminescence imaging was performed thrice weekly using an IVIS 100 imager (PerkinElmer, Waltham, MA). For imaging, mice were injected i.p. with 100 μl of 30 mg/ml luciferin (GoldBio, St. Louis, MO), followed by isoflurane anesthesia and imaging 12 min after luciferin injection (2 min exposure, medium binning).

The cationic liposome polyinosinic-polycytidylic acid adjuvant was prepared in the laboratory as described previously. Using an insulin syringe (BD Biosciences), 50 μl of adjuvant was injected into the right rear footpad of mice while under isoflurane anesthesia. Following injection, mice were immediately treated i.p. with losartan or losartan EXP-3174 metabolite (60 or 10 mg/kg, respectively) and again 24 h later. Animals were euthanized ∼2 h following the second drug treatment, and right popliteal lymph nodes (LNs) were harvested and stored in complete media on ice until processing. Right popliteal LNs harvested from naive, uninjected mice served as controls. LNs were mashed on 40 μM cell strainers using a 3-ml syringe plunger, rinsed with 10 ml complete media, centrifuged at 1200 rpm × g for 5 min, and resuspended in FACS buffer for immunostaining and flow cytometry analysis.

To induce peritonitis, mice (C57BL/6J wild-type, CCR2^−/− and Agtr1^−/−) were injected i.p. with 1 ml thioglycollate as described previously (31). Mice were treated with losartan (60 mg/kg per day i.p.), beginning ∼2 h after thioglycollate injection and continuing once daily until 72 h postinjection. Peritoneal leukocytes were collected at 72 h by flushing the peritoneal cavity with 10 ml PBS, followed by centrifugation and resuspension in ammonium-chloride-potassium solution for red cell lysis. Leukocytes were then washed into FACS buffer, and 4 × 10⁵ cells/well were plated in 96-well round-bottom plates for immunostaining and flow cytometry analysis as previously described. Briefly, nonspecific binding was blocked by adding normal mouse serum (Jackson ImmunoResearch) and unlabeled anti-mouse CD16/32 (eBioscience) to cells before immunostaining. Cells were then incubated with the following panel of directly labeled rat mAbs (eBioscience, San Diego, CA, unless otherwise noted) directed against mouse CD11b (clone M1/70), mouse Ly6C (clone AL-21), mouse Ly6G (clone 1A8), mouse CCR2 (clone 475301; R&D Systems), and mouse F4/80 (clone BM8).

PBMCs were isolated from fresh, EDTA-treated human blood by lysing erythrocytes (×2) with ammonium-chloride-potassium buffer solution (150 mM NH₄Cl, 10 mM KHCO₃, and 0.1 mM Na₂EDTA). PBMCs were washed into serum-free RPMI 1640 (sf-RPMI) and resuspended at 2 × 10⁶ cells/ml. Cultured THP-1 cells were washed into sf-RPMI and resuspended at 6 × 10⁶ cells/ml. Drug stocks (losartan, losartan EXP-3174 metabolite, or RS102895) were diluted to 2× in sf-RPMI. THP-1 cells or PBMCs were diluted 1:1 in media alone (positive and negative controls) or in media containing 2× drug dilutions. Cells were pretreated at 37°C in the incubator for 1 h prior to plating. The chemotactic stimulus for positive control and drug-treated wells consisted of 50 ng/ml recombinant human CCL2 (PeproTech, Rocky Hill, NJ). Negative control wells consisted of sf-RPMI only. THP-1 chemotaxis was conducted in 24-well plates containing 3-μm pore diameter cell culture inserts (Falcon, Corning, NY). For these assays, 600 μl of media, with or without CCL2, was plated in the lower compartment of the plate, whereas 100 μl (3 × 10⁵) THP-1 cells in media, with or without drug, were plated in the upper compartment of the cell culture insert. For PBMC migration assays, 96-well chemotaxis plates (Corning, Corning, NY) with an 8-μm pore diameter were used, and 150 μl of media, with or without CCL2, was plated in the lower compartment of the plate, whereas 50 μl (5 × 10⁴) PBMCs in media, with or without drug, were plated in the upper compartment of the cell culture insert. Cells were allowed to migrate for 4 h. Following migration, nonmigrated cells were removed, wells were washed, and membranes (THP-1 migration) or lower compartment wells (PBMCs) were fixed with 4% paraformaldehyde for 10 min on ice, stained with 3% crystal violet (Sigma-Aldrich, St. Louis, MO), rinsed with dH₂0, and air dried overnight. For analysis of THP-1 chemotaxis, membranes were cut from the cell culture inserts and mounted “migrated-side” up on superfrost plus glass slides using immersion oil. A total of (5) 40× fields per membrane were counted to determine the mean number of monocytes/40× field for each membrane. For PBMC migration assays, 4 × 4-tiled 10× magnification overviews of 96-well plates were obtained for each individual well, and total monocytes per/well were counted using ImageJ (National Institutes of Health).

THP-1 cells (5 × 10⁵ cells/well) were plated in 24-well plates and serum starved overnight for ∼20–24 h in the incubator. The following morning, losartan or losartan EXP-3174 was added to cultures to achieve the indicated treatment concentrations, and cells were pretreated for 2 h prior to CCL2 stimulation. Samples were stimulated with CCL2 (10 nM; PeproTech) for 1 min, quickly pelleted, supernatant discarded, and resuspended in ice-cold lysis buffer (M-PER reagent [Thermo Fisher Scientific, Waltham, MA] containing 1 mM sodium orthovanadate, 100 mM PMSF, 2% SDS, and 1× protease inhibitor mixture [Roche, Basel, Switzerland]) for 10 min on ice. Lysates were then centrifuged at 13,000 rpm × g for 5 min and supernatant removed. For Western analysis (5), a microgram of THP-1 lysate was mixed 1:1 with 2× Laemmli sample buffer containing 5% 2-ME (Bio-Rad Laboratories, Hercules, CA), boiled for 5 min, cooled on ice, and then loaded into a Mini-PROTEAN TGX 4–20% precast polyacrylamide gel (Bio-Rad Laboratories) for electrophoresis (150 V, 45 min). Protein was then transferred to nitrocellulose membranes (95 V, 50 min at 4°C), and membranes were blocked for 1 h at room temperature (RT) with 5% BSA in TBS Tween 20 solution (TBST). After washing in TBST, membranes were incubated with the primary Ab (monoclonal rabbit anti–phospho-p44/42 MAPK, clone D13.14.4E; Cell Signaling Technology, Danvers, MA) diluted in 5% BSA–TBST overnight at 4°C. The following day, membranes were rinsed (×3 with TBST), incubated with the secondary Ab (HRP-linked goat anti-rabbit IgG; Thermo Fisher Scientific), diluted 1:20,000 in 5% BSA/TBST for 1 h at RT. Last, membranes were imaged with chemiluminescent substrate (Clarity Western ECL; Bio-Rad Laboratories) using a ChemiDoc XES⁺ System (Bio-Rad Laboratories).

THP-1 cells (2.5 × 10⁵) in sf-RPMI, with or without losartan or losartan EXP-3174 metabolite at the indicated concentrations, were incubated for 1 h at 37°C in microcentrifuge tubes. Following drug pretreatment, cells were stimulated with 20 nM human rCCL2 for 3 min at 37°C. Immediately following stimulation, the reaction was terminated by the fixation of cells in an equal volume of 4% paraformaldehyde for 15 min at 37°C. Fixed cells were pelleted, washed twice in FACS buffer, and then permeabilized by resuspension in 150 μl of ice-cold 100% methanol for 15 min. Following permeabilization, cells were washed in FACS (×2) and then stained with monoclonal rabbit anti-human phospho-p44/42 MAPK–Alexa Fluor 647 (clone1792G2; Cell Signaling Technology) at 1 μg/ml diluted in FACS for 30 min at RT. Following primary Ab labeling, cells were washed in FACS (×2) and then analyzed by flow cytometry.

Immediately following euthanasia, left lung lobes were dissected, and immersion was fixed in 1% paraformaldehyde-lysine-periodate fixative (1% paraformaldehyde in 0.2 M lysine-HCl, 0.1 M anhydrous dibasic sodium phosphate, and 0.21% sodium periodate [pH 7.4]) for 24 h at 4°C. Following fixation, lungs were placed in a 30% w/v sucrose solution for 24 h at 4°C, prior to embedding and freezing in OCT compound (Tissue-Tek). Embedded tissues were sectioned at 5 μm for immunostaining. Nonspecific binding was blocked by preincubation of sections with 5% donkey serum (Jackson ImmunoResearch, West Grove, PA) in 1% BSA for 30 min at RT. Primary Ab labeling (1:200 anti-GFP, Novus Biologicals; 1:100 anti-F4/80 clone BM8, eBioscience; and 1:100 anti-Ly6C ab76975) was performed at RT for 1 h in 1% BSA. After removal of the primary Ab, tissues were washed with PBS with Tween 20, followed by the addition of the following secondary Abs (diluted 1:200 in PBS with Tween 20) for 30 min at RT: Alexa Fluor 488–conjugated donkey anti-rabbit IgG (GFP), Alexa Fluor 647–conjugated donkey anti-rat IgG (F4/80), and Cy3-conjugated donkey anti-rat IgG (Ly6C) (Jackson ImmunoResearch). Tissues were counterstained with DAPI, cover slipped, and visualized using an Olympus IX83 confocal microscope and Hamamatsu digital camera. Figures were assembled using Adobe Photoshop (CC2016).

For quantification of F4/80⁺ cells within pulmonary micrometastases and CT26-GFP⁺ micrometastasis density within lung sections, both single field and whole slide images were captured using standardized exposure times and an Olympus IX83 disc-spinning confocal fluorescence microscope and Hamamatsu ORCA-R2 digital camera. All image analysis was performed in a blinded fashion using ImageJ software (National Institutes of Health), as described below.

To determine the CT26-GFP⁺ pulmonary micrometastatic burden, whole slide images were taken with a 4× objective (for 40× total magnification) of both the GFP and DAPI channels. Using the DAPI image, a lower threshold was initially determined using a value corresponding to 2 SDs above the mean fluorescence intensity of slide regions devoid of lung tissue. The accuracy of this threshold value to capture the total lung tissue area was ensured visually by a board-certified pathologist and subsequently applied to all images in the dataset to generate masks (outlines) that included only the lung tissue area. CT26-GFP⁺ cells were counted by pixels over a lower threshold limit. This limit was set at a value corresponding to the mean + 2 SDs brightness of control tissue (lung tissue containing no GFP⁺ tumor cells), again visually ensured for quantification accuracy, and this value was universally applied to all images in the dataset. The lung tissue area outlines were then overlaid onto masks of CT26-GFP⁺ cells. CT26-GFP⁺ tumor cell area was recorded as a percentage of total lung tissue area. F4/80⁺ cell infiltration of micrometastases was quantified using single field images captured at 20× (200× total) magnification, with all images being centered on GFP⁺ tumor cell clusters and all micrometastases present in the entire lung section capture per animal. For each field, F480⁺ cells were counted as positive pixels over the lower threshold limit, which was generated in the same manner as described above and corresponded to a value >2 SD above the mean fluorescence intensity of the isotype control and was again visually ensured for accuracy by a board-certified pathologist. This threshold limit was then applied universally to all images in the dataset.

THP-1 cells, human PBMCs, or mouse bone marrow cells were washed and resuspended into sf-RPMI at 8 × 10⁵ cells/ml. Drug stocks (10 mg/ml) were diluted to either 2× or 4× treatment concentrations for either single agent or combination therapy studies, respectively. Then 2 × 10⁵ cells in 250 μl were plated in 24-well plates and diluted either 1:1 with media alone (control) or 2× drug stocks (single agent treatment) or 1:0.5:0.5 (4× drug stocks, combination treatment studies). Cells, with or without drug treatment, were then incubated under standard conditions of 37°C, 5% CO₂, and humidified air for the indicated time periods (1–24 h). Following drug treatment, cells were centrifuged at 1800 rpm × g for 3 min, resuspended in FACS buffer, and stained for CCR2 using a monoclonal mouse anti-human CCR2 Ab (clone TG5/CCR2; BioLegend, San Diego, CA). For human PBMCs, cells were also labeled with mouse anti-human CD14 (clone TUK4; Bio-Rad Laboratories). For mouse bone marrow, cells were stained with the following panel of rat mAbs: mouse CD11b (clone M1/70), mouse Ly6C (clone AL-21), mouse Ly6G (clone 1A8), and mouse CCR2 (clone 475301; R&D Systems). Data were expressed as CCR2 geometric mean fluorescence intensity (gMFI) as the percentage of untreated controls.

Human THP-1 cells were processed and drug treated with losartan, losartan carboxylic acid metabolite (EXP-3174), or a combination of the two drugs at 1 μg/ml for 4 or 24 h, as already described in 2Materials and Methods. RNA was extracted using QIAGEN RNeasy Mini Kit (QIAGEN, Germantown, MD) according to the manufacturer’s instructions. The concentration and purity of the RNA was measured using a NanoDrop 1000 (Thermo Fisher Scientific) with ND-1000 version 3.8.1 software. cDNA was synthesized from 1 μg of RNA using QuantiTect Reverse Transcription Kit (QIAGEN) according to the manufacturer’s instructions. The reaction took place in a MJ Mini Personal Thermal Cycler (Bio-Rad Laboratories). CCR2 expression was measured using RT-PCR with previously published primers (32) and normalized to ACTB (β-actin) (33). Concentrations of 100 nm for the forward primer and 200 nm for the reverse primer were used in the reaction. Master Mix containing SYBR Green dye for florescence indication was used (iQ SYBR Green Supermix; Bio-Rad Laboratories) in a total reaction volume of 10 μl with 20 ng of cDNA. An increase of fluorescence to measure amplification was performed by the Mx3000P (Stratagene-Agilent, Santa Clara, CA) and analyzed using the Mx3000P version 2.0 software. One cycle of 95°C for 10 min followed by 40 cycles of 95°C for 30 s and 60°C for 1 min were used. A dissociation curve cycle was also added to confirm that a single product was being amplified. Data are expressed as fold change relative to untreated control cells.

THP-1 cells (2.5 × 10⁵ cells/well) were plated in 96-well plates in chemokine-labeling buffer (RPMI plus 20 mM HEPES, 10% FBS, 1% l-glutamine, and 1% penicillin/streptomycin) alone (positive control), buffer containing either 30 nM unlabeled human rCCL2 (cold-competition control), losartan, losartan EXP-3174 metabolite, or INCB3284 at the indicated concentrations. Human rCCL2–Alexa Fluor 647 (CAF-2, Almac, Souderton, PA) was then added to all wells to obtain a final concentration of 30 nM. Cells were then incubated for 1 h at 37°C, and ligand binding was subsequently analyzed by flow cytometry. Data were expressed as the percentage of inhibition of ligand binding as determined by the differences in CCR2 gMFI between untreated and drug-treated cells.

The FLIPR Calcium 4 Assay (Molecular Devices) was performed according to the manufacturer’s protocol with minor modifications. HEK293s cells stably transfected with tetO-CCR2 system (pcDNA6-TR, blasticidin resistant and pACMV-tetO-CCR2, Geneticin resistant) were maintained under 700 μg/ml Geneticin (Life Technologies) and 5 μg/ml Blasticidin (Fisher Bioreagents) in complete DMEM plus 10% FBS media. The cells were seeded in poly-d-lysine–coated 96-well black/clear-bottom plates (Becton Dickinson) at 70 K cells per well in 100 μl of DMEM plus 10% FBS plus 2 μg/ml doxycycline. Twenty-four hours after plating, cell culture medium was replaced with 100 μl of Ca Flux buffer (1× HBSS, 20 mM HEPES, and 0.1% BSA), 10 μl of 24× final concentrations of the test compounds in assay buffer, and 110 μl of FLIPR dye (1× dilution; Becton Dickinson). Following at least 1 h incubation at 37°C, the plates were transferred to a FlexStation II 384 Plate Reader. Wells were injected at t = 18 s with 30 μl of 1.6 μM human rCCL2 (generous gift from Tracy Handel laboratory, University of California San Diego) (final in-well concentration of 200 nM), and fluorescence was measured for 150 s, reading every 3 s (Ex485/Em525). The difference between maximum (peak) fluorescence and baseline fluorescence was measured in triplicates for each compound concentration and averaged. The experiments were repeated on at least three different days with the data being normalized to the maximum fluorescence observed on the same day for untreated CCR2-expressing cells.

The CCR2b–Small Bit (SmBiT) and Large Bit (LgBiT)–β-arrestin1–EE plasmids are a generous gift from Asuka Inoue (Graduate School of Pharmaceutical Sciences, Tohoku University, Sendai, Japan). The NanoBiT (34) is a previously published enzyme complementation system derived from the engineered Oplophorus gracilirostris luciferase known as NanoLuc (35). The two fragments of NanoBiT are the SmBiT, which is a variant of the C-terminal fragment of NanoLuc (residues 185-VTGYRLFEEIL-195), and the LgBiT, which is a modification of the remaining larger fragment: 27-VFTLEDFVGD WEQTAAYNLD QVLEQGGVSS LLQNLAVSVT PIQRIVRSGE NALKIDIHVI and IPYEGLSADQ MAQIEEVFKV VYPVDDHHFK VILPYGTLVI DGVTPNMLNY FGRPYEGIAV-184.

Both fragments have been mutated to reduce their mutual binding affinity as described by Dixon et al. (34). CCR2b-SmBiT was obtained by fusing the NanoBiT SmBiT, with a flexible 15-AA linker (GGSGGGGSGGSSSGG) preceding it, to the C terminus of human CCR2b in pCAGGS. For this, CCR2b open reading frame was PCR amplified using two oligonucleotides (5′-AGAATTGAGCTCCCGGGTACCGCCACCATGCTGTCCACATC-3′ and 5′-GGACAAAGAAGGAGCCGGGGGATCTGGGGGGGGG-3′) and inserted into a pCAGGS plasmid vector containing the linker and SmBiT by using an NEBuilder HiFi DNA Assembly System (New England BioLabs). LgBiT–β-arrestin1–EE (36) was obtained by fusing the NanoBiT LgBiT to the N terminus of clathrin-binding–deficient variant of human β-arrestin1 incorporated in pCAGGS vector with a flexible 16-AA linker GGSGGGGSGGSSSGGT between the two. The EE variant contains two mutations (R393E, R395E) in the clathrin/AP-2–binding motif of β-arrestin1, which leads to enhanced retention at the cell plasma membrane and hence an increased receptor recruitment signal (37). The constructs were propagated in Escherichia coli using ampicillin (100 μg/ml) as a bacterial selection marker.

HEK293t cells were plated in a 6-cm dish. Following a 24 h incubation at 37°C in 5% CO₂, cells were transiently transfected with CCR2b-SmBiT and LgBiT–β-arrestin1–EE plasmids (3 μg of each DNA per 6-cm dish). Twenty-four hours after transfection, the cells were lifted with PBS containing 0.2 mM EDTA, centrifuged for 5 min at 400 × g, resuspended in assay buffer (1× HBSS, 5 mM HEPES [pH 7.2], and 0.05% BSA), and normalized to 1.2 × 10⁶ cells/ml. CTZ-n (no. 501216836, from 5 mM stock in ethanol; Thermo Fisher Scientific) was added to the cell suspension to achieve a final concentration of 10 μM. After that, 80 μl of cell suspension was transferred to a 96-well black/clear-bottom plate (no. 353219 Falcon). The plate was incubated at RT for ∼90 min and protected from light. Ten microliters of assay buffer or ten microliters of 10× final concentrations of the test compounds (prepared from 10 mM DMSO stocks and diluted in assay buffer) were added to the wells as per plate map, incubated for 10 min at RT, and protected from light. A backing tape (no. 6005199; PerkinElmer) was applied to the bottom of the plate, after which the base luminescence was read for each well using PerkinElmer VICTOR X Light 2030 (1 s, no filter). Next, 10 μl of 2 μM CCL2 in assay buffer was added to each well to a final concentration of 200 nM. The cells were incubated at RT for 10 min and protected from light, after which the plate was read again for end point luminescence. The results were analyzed with GraphPad Prism version 7.0b.

All data expressed as means ± SD unless otherwise noted. Statistical significance was determined by a two-tailed, unpaired Student t test or one-way ANOVA with Tukey posttest for multiple group comparisons. All statistical analyses were performed using GraphPad Prism software (La Jolla, CA).

All animals were housed in microisolator cages in the laboratory animal facility at Colorado State University, and all animal procedures were approved by the Institutional Animal Care and Use Committee at Colorado State University.

The effects of losartan and the primary carboxylic acid metabolite (losartan carboxylic acid; EXP-3174) on CCL2-stimulated migration of human monocytic cells (THP-1) were investigated using a trans-well chemotaxis assay (Fig. 1A). Treatment with losartan or EXP-3174 significantly inhibited CCL2-stimulated THP-1 migration by up to 90% compared with untreated cells (*p < 0.05). Furthermore, the magnitude of monocyte migration inhibition was comparable to that observed for the specific small molecule CCR2 competitive antagonist RS102895, when evaluated at equimolar concentrations, and the significant inhibition of CCL2-mediated THP-1 chemotaxis was observed at losartan treatment concentrations as low as 100 ng/ml (Supplemental Fig. 1A). Importantly, this blockade of monocyte migration was not secondary to primary cytotoxic effects of losartan or EXP-3174 on THP-1 cells as treatment concentrations up to 100 μg/ml had no effect on THP-1 growth or survival (Supplemental Fig. 1B). Additionally, in contrast to the results of the CCL2 chemotaxis assays, losartan treatment did not significantly block THP-1 migration to the chemokine SDF-1α, a known potent THP-1 chemoattractant and GPCR agonist (38), suggesting that losartan’s inhibitory effects were specific to CCL2-mediated chemotaxis (Supplemental Fig. 1C). The ability of losartan to block the migration of primary human monocytes was also assessed using human PBMCs (Fig. 1B, 1C). In this study, losartan again significantly inhibited monocyte migration (by ∼50%; *p < 0.04) at 1 μg/ml, a concentration approximately equal to the maximum plasma levels observed for a single 100 mg oral dose of losartan, a dose routinely used for the treatment of hypertension (39).

Next, the ability of losartan and EXP-3174 to inhibit in vivo IM recruitment was assessed using a murine footpad inflammation model, which we have used previously to assess CCL2–CCR2-dependent monocyte recruitment to draining LNs (Supplemental Fig. 1D) (40). Using this model, we found that the administration of losartan or EXP-3174 significantly reduced the recruitment of IMs to draining LNs by over 75% (**p < 0.01) (Fig. 1D, 1E, Supplemental Fig. 1E).

In a second approach, the thioglycollate model of aseptic peritonitis was used to evaluate losartan effects on monocyte recruitment to the peritoneal cavity. This model has been previously used for a preclinical evaluation of other small molecule chemokine receptor antagonists (41, 42). A key feature of this model is that monocyte accumulation in the peritoneal cavity is primarily dependent on CCL2–CCR2 signaling (43, 44). In the current study, we found that losartan treatment (60 mg/kg per day i.p.) significantly reduced the percentage of F4/80⁺/CCR2⁺ monocytes accumulating in the peritoneal cavity at 72 h after thioglycollate injection to a level equivalent to that observed in CCR2^−/− mice (****p < 0.0001) (Fig. 1F, 1J).

Given the potent monocyte migration inhibitory activity of losartan, additional experiments were conducted to better elucidate the molecular interaction between losartan/EXP-3174 and CCR2. Using the THP-1 cell line (which expresses high levels of CCR2), the effects of losartan and EXP-3174 metabolite on CCL2 binding and CCL2-induced intracellular Ca²⁺ mobilization were studied. The specific CCR2 orthosteric antagonists BMS-681 and INCB3284 were used as positive controls (45, 46). Surprisingly, ligand-binding studies demonstrated that both losartan and EXP-3174 completely failed to block CCL2 binding to CCR2 on THP-1 cells at concentrations ranging from 1 nM to 100 μM (Fig. 2A–C).

In addition, losartan and EXP-3174 had only modest inhibitory effects on CCL2-stimulated cytosolic calcium release and, at equimolar concentrations, were significantly less potent than BMS-681 (Fig. 2D). Pretreatment with losartan and EXP-3174 resulted in a mean maximal inhibition of CCL2-induced calcium responses by 22 and 53%, respectively, at doses of 50 μM (**p < 0.01) (Fig. 2D, 2E), a concentration equivalent to the maximum plasma concentration (C_max) observed in our mouse studies. Significant inhibition of CCL2-induced calcium release was observed with EXP-3174 treatment at concentrations as low as 80 nM, although still significantly less potent than BMS-681 (Fig. 2E).

Losartan and EXP-3174 were also evaluated in a second CCR2 functional assay assessing the drugs’ ability to block CCL2-induced β-arrestin recruitment to CCR2. Consistent with results of the calcium flux assays, the significant inhibition of β-arrestin recruitment to CCR2 was again only observed at the high, C_max-equivalent losartan dose of 50 μM (Fig. 2F–H). Thus, losartan failed to block CCL2 binding to CCR2 and, consistent with this observation, only slightly reduced CCL2-induced Ca²⁺ release and β-arrestin recruitment, whereas pure orthosteric CCR2 antagonists were highly active at equivalent concentrations.

The preceding studies indicated that losartan blockade of monocyte migration was not mediated by the competitive inhibition of CCL2 binding to CCR2 nor to any downstream effects on CCL2-mediated cytosolic Ca²⁺ release. Therefore, studies were done next to determine whether losartan inhibited additional targets in the CCL2–CCR2 signaling pathway. Prior studies have demonstrated that CCL2-induced integrin activation and CCR2-dependent monocyte chemotaxis is mediated through the MAPK/ERK (MAPK/ERK) pathway (47–50). Furthermore, these studies also evaluated the specific second messengers involved in CCL2–CCR2 signaling and demonstrated that CCL2-induced activation of ERK1/2 was in fact independent of changes in cytosolic [Ca²⁺] or β-arrestin–mediated receptor internalization and signaling (47).Thus, the ability of losartan to inhibit CCL2-induced ERK1/2 activation was evaluated in THP-1 cells using Western blot and intracellular flow cytometry assays to assess ERK1/2 phosphorylation (Fig. 3). THP-1 cells were pretreated with losartan and then stimulated with 20 nM CCL2, a concentration previously reported to induce strong and rapid activation (phosphorylation) of ERK1/2 (47). Both losartan and EXP-3174 significantly inhibited ERK activation in response to CCL2, reducing phospho-ERK1/2 mean fluorescence intensity to levels ∼30 and 11%, respectively, of those observed in untreated, CCL2-stimulated positive control cells (Fig. 3A, 3B). Similarly, ERK1/2 phosphorylation following either acute (Fig. 3C) or chronic/prolonged (Fig. 3D) CCL2 stimulation (as would be expected in a tumor-bearing individual) also revealed that both losartan and EXP-3174 inhibited ERK activation in response to CCL2. For example, losartan and EXP-3174 treatment reduced pERK levels by 29 and 42% following acute CCL2 stimulation (Fig. 3C), respectively, and by 67 and 65% following prolonged (24 h) CCL2 stimulation, respectively (Fig. 3D). A dose-dependent inhibition of ERK1/2 phosphorylation in human CD14⁺ monocytes by both losartan and EXP-3174, as assessed by flow cytometry, was also observed upon ex vivo treatment of PBMCs (Supplemental Fig. 2A, 2B).

The observed effects of losartan and EXP-3174 on CCL2-induced ERK activation and monocyte chemotaxis could have been mediated by drug-induced receptor downregulation. To address this possibility, the effects of losartan or EXP-3174 pretreatment on CCR2 expression by THP-1 cells was assessed by flow cytometry. Results of these experiments demonstrated a modest dose and time-dependent reduction in CCR2 expression, occurring as early as 4 h posttreatment (Fig. 3E), with additional receptor downregulation occurring for up to 24 h after losartan treatment (Fig. 3F). The CCR2 downregulation effect was additive when losartan and EXP-3174 were combined (Fig. 3E, 3F). In addition, decreased CCR2 cell surface expression was also observed in human peripheral blood CD14⁺ monocytes (Supplemental Fig. 2C) and murine bone marrow–derived CD11b⁺/Ly6C⁺ monocytes following losartan treatment (Supplemental Fig. 2D).

To identify the potential mechanism behind the losartan-mediated reduction in surface CCR2 expression, we evaluated the effects of losartan, the EXP-3174 metabolite, or a combination drug treatment on CCR2 mRNA expression in THP-1 cells. These experiments were conducted at the same treatment concentrations and time points performed for the flow cytometric evaluation of CCR2 surface expression shown in Fig. 3E and 3F. Results of these experiments demonstrated no effect of losartan treatment on the level of CCR2 gene expression in THP-1 cells (Supplemental Fig. 2E, 2F), ruling out the downregulation of CCR2 transcription as a possible mechanism for the reduction in cell surface CCR2 protein expression.

Immunofluorescent labeling and flow cytometric evaluation of surface expression of CCR2 or other chemokine receptors is frequently used as a surrogate assay to quantify drug or ligand-induced chemokine receptor internalization (51–53). Thus, it was concluded that losartan induced moderate downregulation of CCR2 surface expression on monocytes, independent of an effect on CCR2 gene expression, via a mechanism that most likely involved the induction of receptor internalization.

Based on these data, two plausible mechanisms for losartan inhibition of monocyte migration were apparent: 1) inhibition of CCL2-induced ERK signaling and 2) CCR2 receptor downregulation. We propose that of the two mechanisms, the inhibition of ERK signaling was likely the most important pharmacodynamic effect of losartan on monocytes, given that ERK1/2 activation is essential in regulating monocyte migration in response to CCL2. Supporting this notion, ex vivo inhibition of CCL2-induced ERK phosphorylation is a primary pharmacodynamic end point for the Pfizer small molecule CCR2 antagonist (PF-04136309) currently being evaluated in a phase II clinical trial (ClinicalTrials.gov, NCT02732938). Thus, our results were most consistent with a model wherein the primary effects of losartan and EXP-3174 were mediated by both the direct inhibition of CCR2 signaling through the blockade of ERK activation as well as a contribution from the downregulation of cell surface CCR2 receptor expression.

Although the preceding studies suggested a blockade of CCL2–CCR2–ERK signaling was the primary pharmacodynamic effect of losartan on monocytes, it remained formally possible that this inhibitory activity may have been mediated indirectly via the engagement of the AT1R, the biological target for losartan antihypertensive activity. For example, G-protein coupled receptor cross-talk through mechanisms including receptor heterodimerization and allosteric trans-inhibition have been previously described, including examples involving CCR2 (54–57). To address this issue directly, additional studies were performed using AT1R^−/− mice.

Importantly, we first noted that the lack of AT1R expression had no demonstrable effect on CCL2-mediated monocyte migration, either in vitro or in the thioglycollate peritonitis model (Fig. 4B), thereby eliminating any essential role for angiotensin II–AT1R signaling in these inflammation models. Nonetheless, we observed that ex vivo treatment of AT1R^−/− bone marrow cells with losartan still significantly blocked CCL2-mediated monocyte chemotaxis (Fig. 4A). Moreover, the treatment of AT1R^−/− mice with losartan (60 mg/kg per day i.p.) significantly reduced the accumulation of IMs in the peritoneal cavity to a degree equivalent to CCR2^−/− mice (****p < 0.0001) (Fig. 4B). Furthermore, the inhibitory effects of losartan on CCL2-induced ERK phosphorylation were maintained in AT1R^−/− mice, along with the drug’s effects on the downregulation of cell surface CCR2 expression. For example, the treatment with losartan significantly reduced CCL2-mediated ERK1/2 phosphorylation in Ly6G⁻/CD11b⁺/Ly6C^Hi bone marrow monocytes of AT1R ^−/− mice by ∼50% (Fig. 4C, 4D). Additionally, once daily, losartan treatment significantly reduced the cell surface expression of CCR2 on AT1R knockout (KO) peritoneal macrophages collected 72 h after thioglycollate injection (Fig. 4E, 4F).

These results therefore suggested that losartan interruption of CCR2 signaling was not mediated by AT1R signaling and also ruled out mechanisms such as AT1R–CCR2 heterodimerization and/or allosteric trans-inhibition. Thus, losartan remained fully active as a monocyte migration blocking agent and inhibitor or ERK1/2 signaling in response to CCL2 even in the absence of the primary losartan-intended receptor AT1R (Fig. 4).

The preceding studies suggested that losartan may have a potential use as an antimetastatic agent, given the recognized importance of IMs in the early metastatic events (10). Previous studies have documented losartan-induced suppression of the growth of primary tumors, but to date, the effects of losartan or other ARBs on tumor metastasis have not been studied (29, 30, 58). Therefore, studies were done to assess the impact of losartan treatment on metastasis-induced monocyte recruitment to the lungs, using the CT26 lung metastasis model. In this model, tumor metastasis produced a significant increase (∼4-fold) in the recruitment of CD11b⁺/Ly6C^Hi IMs to the lungs (Fig. 5A, 5B, 5E) within 72 h of tumor cell injection. In addition, the use of CCR2^−/− mice in this model demonstrated that the observed monocyte recruitment was entirely dependent on the monocyte expression of CCR2 (Fig. 5D, 5E).

Mice with lung metastases were treated with losartan at a dose of 60 mg/kg per day i.p. beginning 24 h after the tumor cell injection and were sacrificed 72 h later. The numbers of infiltrating monocytes were quantitated by flow cytometry and immunofluorescence imaging. IMs were identified as Ly6G⁻/SiglecF⁻/CD11c⁻/CD11b⁺/Ly6C^Hi cells, and MAMs were enumerated as CD11b⁺/F4/80⁺ cells. Losartan treatment significantly reduced the percentages of IMs and MAMs in the lung following tumor injection by 70 and 36%, respectively, compared with untreated animals. The reduction in the numbers of monocytes and macrophages was roughly equivalent to the reduction observed in CCR2^−/− mice (Fig. 5C, 5E, 5F, 5G).

We also noted a striking reduction in the number of CT26-GFP⁺ micrometastatic colonies in the lungs of losartan-treated mice (Fig. 5H). For example, the area occupied by CT26-GFP micrometastases, quantified as a percentage of the total evaluated lung lobe area, was reduced by 70% in losartan-treated mice (and by 90% in CCR2^−/− mice) compared with untreated control animals (Fig. 5I). These results suggested that losartan treatment and the blockade of early tumor-mediated monocyte recruitment to the lung was also associated with decreased tumor cell growth during the early, postcolonization time period.

Studies were done next to determine whether losartan treatment could produce sustained inhibition of CCR2 signaling, monocyte recruitment, and the suppression of tumor metastasis growth. Luciferase-expressing 4T1 breast or CT26 colon carcinoma cells were injected i.v., losartan treatment (60 mg/kg per day i.p.) was initiated 24 h postinjection, and the lung metastatic burden was monitored three times weekly using bioluminescence imaging. Daily treatment with losartan significantly reduced both CT26 and 4T1 pulmonary metastatic burden by 64 and 90%, respectively, as quantified by bioluminescent imaging (CT26: Fig. 6A–C, and 4T1: Fig. 7A, 7B). In the 4T1 model, the losartan-mediated reduction in the metastatic tumor burden significantly prolonged the overall survival (Fig. 7C, *p = 0.04). The reduction in lung metastasis was confirmed via histopathological evaluation of the lungs at euthanasia (Fig. 7D).

Importantly, in both the CT26 and 4T1 models, daily losartan treatment resulted in a significant, sustained inhibition of CD11b⁺/Ly6C^Hi monocyte recruitment to the lungs of metastasis-bearing mice as revealed by flow cytometric analysis. For example, at study termination (day 19 for CT26 mice and day 14 for 4T1 mice), there was a 2-fold reduction in the percentage of lung monocytes in losartan-treated mice as compared with vehicle (saline)-treated mice (Figs. 6D, 6E, 7E, 7F). Immunofluorescent staining of tissue sections of CT26 and 4T1 pulmonary metastases confirmed the reduction in tumor-infiltrating F4/80⁺ and CD11b⁺ myeloid cells, respectively (Figs. 6D, 7E).

Monocytes are known to be a rich source of VEGF production, and therefore microvessel density in CT26 tumors was assessed via CD31 immunofluorescence imaging. There was a significant 35% reduction in tumor microvessel density in losartan-treated mice as compared with untreated tumor-bearing mice (Fig. 6F). Taken together, these results demonstrated that daily losartan treatment effectively suppressed breast and colon carcinoma pulmonary colonization and growth, an effect that was associated with a significant reduction in the number of lung monocytes and tumor-associated angiogenesis.

Previous studies have demonstrated that angiotensin II–AT1R signaling within the tumor stroma can drive tumor-promoting inflammation (54, 59) as well as tumor angiogenesis (55). Thus, it was plausible that the observed antitumor effects of losartan in our metastasis models may have been mediated by the direct inhibition of AT1R signaling, independent of the observed blockade on monocyte and tumor-macrophage recruitment. To address this question, the experimental metastasis assays and losartan treatments were done in BALB/c CCR2^−/− mice. We hypothesized that if the antimetastatic effects of losartan were mediated in part through AT1R blockade, we should observe enhanced suppression of CT26 metastasis growth in losartan-treated CCR2^−/− mice. Although metastasis was delayed in CCR2^−/− mice (Supplemental Fig. 3A, 3B), losartan treatment did not exert an additive effect in suppressing metastasis in these animals (Supplemental Fig. 3A, 3B), suggesting that the presence of CCR2 was essential and necessary for losartan antitumor activity.

In addition, we quantified CCL2 and angiotensin II production by tumor cells in vitro as well as in vivo serum angiotensin II concentrations in mice with CT26 metastases. Both 4T1 and CT26 cells produced substantially more CCL2 than angiotensin II (Supplemental Fig. 3C) and significantly less angiotensin II compared with the positive control cell line Lewis lung carcinoma (56) (Supplemental Fig. 3D). Furthermore, serum angiotensin II concentrations were not elevated in CT26 metastasis-bearing control (175.7 ± 46.8) or losartan-treated mice (143.4 ± 37.4) compared with mice without tumors (176.1 ± 9.5) (Supplemental Fig. 3E, mean ± SEM pg/ml).

Last, 72 h treatment of CT26 or 4T1 cells with a losartan concentration roughly equivalent to the overall exposure observed in our in vivo pharmacokinetic studies (as determined by the area under the curve [AUC_0–∞]) revealed that there was no direct effect of losartan on tumor cell survival or proliferation (Supplemental Fig. 3F). Therefore, these studies excluded angiotensin II–dependent losartan antitumor activity and ruled out direct tumor cell cytotoxic or antiproliferative action of losartan via AT1R inhibition.

Finally, pharmacokinetic analysis after 14 d of i.p. losartan dosing in mice was performed to address the following two questions: 1) Were plasma losartan concentrations in treated mice equivalent to concentrations used in our in vitro studies, and 2) Were the losartan doses used in our mouse experimental metastasis studies relevant to drug concentrations achieved in humans treated with currently recommended antihypertension doses of losartan? With respect to the first question, Supplemental Fig. 4A presents the mean plasma concentrations of both losartan and EXP-3174 in mice following a single i.p. dose of 60 mg/kg on day 14, and Table I in Supplemental Fig. 4B summarizes pertinent pharmacokinetic parameters. Indeed, the C_max and overall exposure (AUC_0–∝) in these animals was well within the range of the demonstrated effective concentrations in our in vitro CCR2 functional assays. For example, the mean maximal plasma concentration of losartan and EXP-3174 in these mice was 24 and 23 μg/ml, respectively, which is substantially greater than the 10 μg/ml concentration that demonstrated the maximal in vitro inhibition of CCL2-induced ERK1/2 phosphorylation in THP-1 cells in our in vitro studies.

For in vivo dosing, we found that the overall exposure to losartan in treated mice (AUC_0–∝ 13 μg/h/mL) was ∼5–6 times the drug concentrations observed in humans administered maximum losartan doses for hypertension or for treatment of Marfan syndrome (39, 57). Although these data suggest that dose-escalation studies of losartan in humans may be needed to achieve comparable CCR2 pharmacodynamic end points, it should be noted that a concentration of 1 μg/ml losartan significantly inhibited human CCL2 monocyte migration in vitro, suggesting that the high losartan concentrations achieved in mice may not be required for full activity in humans.

IMs promote multiple steps of the metastatic cascade (9). Multiple studies have demonstrated a critical role for CCL2–CCR2 signaling in regulating monocyte recruitment to metastases (10, 60–62). Clinically, tumor monocyte density, numbers of circulating monocytes or CCL2 concentrations, and tumor CCL2 expression are all well known as predictors of prognosis of various human malignancies (15–17, 63). Thus, the CCL2–CCR2 axis has become an important potential target for tumor immunotherapy. For example, there are currently two recently completed trials of CCR2-targeted therapies in cancer patients (ClinicalTrials.gov, NCT01015560 and NCT02732938). However, these trials and previous trials targeting CCL2 for cancer immunotherapy have thus far been unsuccessful in achieving their primary study end points (19). The reasons for these trial failures are not fully understood, but one explanation has been a failure to fully block CCR2 signaling or to fully neutralize circulating CCL2. Thus, there remains an opportunity for additional novel approaches to more completely block the CCL2–CCR2 signaling axis for cancer immunotherapy.

The ability of the ARB drug losartan to potently block monocyte migration and monocyte-mediated inflammatory responses opened the possibility that this drug might be repurposed as a CCR2 antagonist for the prevention or early treatment of cancer metastasis, particularly in combination therapy protocols. Our interest in losartan was prompted in part by our frustrating inability to consistently suppress tumor growth using known pure CCR2 antagonist drugs (S. Dow, A. Guth, and D. Regan, unpublished observations). Prior studies have reported losartan-induced antitumor activity, but the effects of losartan on myeloid cell responses were not examined in these studies (29, 30). Therefore, in the current study, we investigated alternative explanations for how losartan might inhibit tumor growth and focused on a possible role of the blockade of monocyte recruitment through the inhibition of CCR2 signaling.

In our studies, in vitro chemotaxis assays and in vivo models of acute inflammation and experimental pulmonary metastasis both demonstrated that losartan and its primary metabolite (EXP-3174) effectively inhibited CCL2–CCR2-mediated IM recruitment. Previous studies of losartan and other ARBs in mouse models of atherosclerosis and immune-mediated encephalomyelitis have demonstrated immunomodulation of monocyte and macrophage activity by ARBs, although in those studies, it was concluded that the primary mechanism of action involved the inhibition of angiotensin II–AT1R–regulated inflammation. However, our results now suggest that losartan functions primarily as an inhibitor of CCL2–CCR2-mediated inflammation. Indeed, our findings indicate that the inhibition of AT1R signaling plays little or no role in the anti-inflammatory effects of losartan treatment. Specifically, our studies in AT1R^−/− mice, wherein the suppressive effects of losartan on CCL2-directed monocyte migration, CCR2 cell surface expression, and CCL2-induced ERK phosphorylation remained fully active both in vitro and in vivo, are consistent with a model in which AT1R signaling plays no role in regulating the effects of losartan on monocyte migration.

Instead, we provide evidence using three distinct chemokine receptor function assays that losartan and its primary EXP-3174 metabolite functionally antagonize CCL2–CCR2 signaling independentof any signaling contribution or interaction with its known AT1R target. Our results suggest that the primary pharmacodynamic effect of losartan on monocyte activity occurs at the level of ERK phosphorylation. Highlighting the significance of this finding and substantiating ERK phosphorylation as an important measure of CCR2 target engagement, ex vivo inhibition of CCL2-induced ERK phosphorylation is a secondary outcome measure in a recent clinical trial of the Pfizer small molecule CCR2 antagonist PF-04136309 (ClinicalTrials.gov, NCT02732938). In our studies, losartan inhibition of CCL2-induced ERK phosphorylation occurred independent of the blockade of CCL2 ligand binding. Overall, these data are most consistent with a mechanism of noncompetitive antagonism of CCR2 signaling. Interestingly, recent crystallography and molecular pharmacology studies of CCR2 have described the presence of a novel, intracellular, allosteric binding site for certain small molecule, noncompetitive CCR2 antagonists (46, 64). The observed lack of losartan inhibition of CCL2 ligand binding is, however, different from the data reported for these novel allosteric antagonists of CCR2 (64).

Thus, the overall results of our CCR2 functional studies and in vivo monocyte recruitment assays in AT1R-deficient mice suggest the presence of an alternative yet unknown mechanism of the losartan-functional antagonism of CCR2. Although the inhibitory effect of losartan on CCL2-induced ERK phosphorylation could feasibly occur at any point in the ERK signaling cascade downstream of CCR2, the inability of losartan to suppress SDF-1α–mediated chemotaxis suggests a more upstream effect of the drug at the level of CCR2 itself. Similar to CCL2, SDF-1α is also a potent chemoattractant of THP-1 cells whose chemotactic effects have been shown to be dependent on ERK signaling. Thus, if losartan nonspecifically inhibited ERK signaling at any level downstream of the CCR2 receptor, we would have expected some degree of inhibition of SDF-1 chemotaxis in THP-1 cells, which was not observed in our studies (38, 65, 66). Nevertheless, these data do not fully elucidate the specific activity of losartan on CCR2 signaling, and additional experiments to assess losartan effects on CCR2 activation, specifically CCL2-induced ERK phosphorylation, following site-directed mutagenesis of both the ortho- and allosteric binding pockets of CCR2, are currently underway.

Losartan was also investigated in this study for its use as a repurposed antimetastatic drug. The effects of losartan treatment on both early and later events in the metastatic cascade were investigated. We found that early tumor colonization of the lung was strongly associated with the recruitment of Ly6C^Hi monocytes in a CCR2-dependent manner. Losartan therapy prevented early (72 h) tumor cell colonization and subsequent monocyte recruitment and MAM accumulation to a degree similar to that observed in CCR2^−/− mice. Longer term daily losartan treatment also suppressed experimental metastasis growth in two different tumor models and was associated with the sustained inhibition of monocyte recruitment and metastasis-associated myeloid cells.

Previous studies in mice have also demonstrated antitumor activity by losartan, albeit not in metastasis models, by using doses three to five times greater than those used in our studies (29, 30, 58). Recent retrospective analyses of clinical data for patients being treated for hypertension have shown a correlation between the use of losartan, or other ARBs and ACE inhibitors, and improved outcomes in patients with pancreatic, breast, or lung cancer (67–70). In the preclinical rodent studies, losartan’s antitumor activity was attributed to antiangiogenic or anti–TGF-β signaling effects as a result of the primary inhibition of angiotensin II–AT1R signaling. However, it should be noted that many of these previously described losartan effects (e.g., inhibition of tumor angiogenesis, tumor TGF-β production) could also be readily explained by inhibiting monocyte recruitment as macrophages are known to stimulate both tumor angiogenesis and TGF-β production (71, 72).

Losartan has a long record as a safe antihypertensive drug and could therefore be rapidly repurposed as a novel immunotherapeutic drug. One key issue in repurposing losartan is to establish pharmacokinetic equivalency. Therefore, we compared the pharmacokinetics of high-dose losartan administered to mice with previously published losartan pharmacokinetic studies in humans. The direct comparison of losartan C_max and AUC_0–∞ observed in mice in our study suggested that the 60 mg/kg per day dose used in mice resulted in ∼6-fold higher drug levels compared with those observed for losartan doses typically used in humans for the treatment of hypertension. However, these drug concentrations may be much higher than those required to effectively block CCR2 signaling. For example, the losartan concentrations that exhibited the suppression of chemotaxis and in vitro CCR2 functional assays in our studies were within the range of the C_max and AUC_0–∞ observed for a single 200 mg oral dose in prior pharmacokinetic studies in humans (39). Thus, losartan doses in the range of 2–3 mg/kg should achieve these target plasma concentrations. There is precedent for using higher doses of losartan safely as in the example of Marfan syndrome patients treated with high-dose losartan, which was safe and well tolerated (57, 73).

Losartan is a Food and Drug Administration–approved drug with a long safety record and high therapeutic index (74, 75). Our studies demonstrate a unique and previously undescribed mechanism of functional antagonism of CCL2–CCR2 signaling and monocyte recruitment by losartan and its primary EXP-3174 metabolite. These studies also show that daily losartan therapy is effective in suppressing experimental metastasis growth, associated with a sustained blockade of IM mobilization and the accumulation of MAMs. Overall, these findings provide evidence for an important new off-target pharmacological effect of losartan, in addition to its previously reported effects on TGF-β signaling. Indeed, most or all of the previously reported anti-inflammatory effects of losartan treatment can be explained by monocyte migration inhibition. Thus, further clinical investigations of losartan for the modulation of the TME and tumor immunity are warranted.

We thank Dr. Tracy Handel (University of California San Diego) for valuable input regarding CCR2 functional assays. Additionally, we thank Dr. Daniel Gustafson, Dr. Ryan Hansen, and Paul Lunghofer (Flint Animal Cancer Center, Colorado State University) for conducting the losartan pharmacokinetic studies in mice. The authors also thank Asuka Inoue (Graduate School of Pharmaceutical Sciences, Tohoku University, Sendai, Japan) for providing the CCR2b-SmBiT and LgBiT–β-arrestin1–EE plasmids.

The authors have no financial conflicts of interest.