A ROS-Responsive Aspirin Polymeric Prodrug for Modulation of Tumor Microenvironment and Cancer Immunotherapy

Abstract

Open AccessCCS ChemistryRESEARCH ARTICLE1 Dec 2020A ROS-Responsive Aspirin Polymeric Prodrug for Modulation of Tumor Microenvironment and Cancer Immunotherapy Sheng Ma, Wantong Song, Yudi Xu, Xinghui Si, Yu Zhang, Zhaohui Tang and Xuesi Chen Sheng Ma Key Laboratory of Polymer Ecomaterials, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022 University of Science and Technology of China, Hefei 230026 Jilin Biomedical Polymers Engineering Laboratory, Changchun 130022 , Wantong Song *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] Key Laboratory of Polymer Ecomaterials, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022 Jilin Biomedical Polymers Engineering Laboratory, Changchun 130022 State Key Laboratory of Molecular Engineering of Polymers (Fudan University), Shanghai 200433 , Yudi Xu Key Laboratory of Polymer Ecomaterials, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022 Jilin Biomedical Polymers Engineering Laboratory, Changchun 130022 University of Chinese Academy of Sciences, Beijing 100039. , Xinghui Si Key Laboratory of Polymer Ecomaterials, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022 University of Science and Technology of China, Hefei 230026 Jilin Biomedical Polymers Engineering Laboratory, Changchun 130022 , Yu Zhang Key Laboratory of Polymer Ecomaterials, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022 Jilin Biomedical Polymers Engineering Laboratory, Changchun 130022 , Zhaohui Tang Key Laboratory of Polymer Ecomaterials, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022 University of Science and Technology of China, Hefei 230026 Jilin Biomedical Polymers Engineering Laboratory, Changchun 130022 and Xuesi Chen *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] Key Laboratory of Polymer Ecomaterials, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022 University of Science and Technology of China, Hefei 230026 Jilin Biomedical Polymers Engineering Laboratory, Changchun 130022 https://doi.org/10.31635/ccschem.020.202000140 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesTrack Citations ShareFacebookTwitterLinked InEmail Tumor-promoting inflammation is accompanied by cancer initiation, progression, and metastasis. Cyclooxygenase-2 (COX-2) and its downstream product, prostaglandin E2 (PGE2), play critical roles in tumor-promoting inflammation. Several studies have revealed the potential of COX-2 inhibition in improving cancer response to chemotherapy, as well as immunotherapy. Aspirin, a nonsteroidal anti-inflammatory drug, has been reported as a COX-2 inhibitor. However, as a small molecule drug with a carboxyl group, there is still the lack of effective methods of preparing polymer–aspirin conjugates with tumor stimuli-responsive release properties. Herein, we synthesized a reactive oxygen species (ROS)-responsive aspirin polymeric prodrug (P3C-Asp) via Passerini three-component reaction between aspirin, 4-formylbenzeneboronic acid pinacol ester, and 5-isocyanopent-1-yne, followed by copper (I)-catalyzed alkyne-azide cycloaddition "click" reaction of the aspirin prodrug with dextran (DEX). The P3C-Asp could release aspirin and salicylic acid in response to tumor-specific stimuli. In the murine colorectal cancer model, P3C-Asp suppressed tumor growth effectively without significant side effects and eradicated tumors when combined with the immune checkpoint inhibitor, anti-PD-1 antibody (aPD-1). Further analysis revealed that the suppression was attributable to changes in the immune microenvironment, including reduced PGE2 content, as well as increased infiltration of CD8+ T cells and M1 macrophages. The results mentioned above proved that targeting COX-2 pathway with a proper polymeric prodrug might be a useful strategy for cancer immunotherapy. Download figure Download PowerPoint Introduction Tumor-promoting inflammation is one of the hallmarks of cancer.1,2 Inflammation, especially chronic inflammation, is thought to be one of the most important environmental factors influencing tumorigenesis, progression, metastasis, as well as immune surveillance.3 The tumor microenvironment (TME) is composed of cancer cells, immune cells, and surrounding stromal cells.4 Due to the high degree of cell plasticity of immune cells, TME reprograms specific immune cells into protumor phenotypes such as myeloid-derived suppressor cells (MDSCs), which promote tumor progression and immune escape.5 MDSCs are derived from immature myeloid cells (IMCs), which are precursors of dendritic cells, macrophages, and granulocytes. In general, IMCs differentiate into mature cells with the possession of immune function.6 However, under TME condition, the maturation of these IMCs is blocked by cyclooxygenase-2 (COX-2) and prostaglandin E2 (PGE2) secreted by tumor cells, and these IMCs become immunosuppressive MDSCs.7 The MDSCs, in turn, suppress the activation and function of CD8+ T cells by depredating l-arginine and generating reactive oxygen species (ROS), promote regular T cells (Tregs) by secreting TGF-β, shift M1 macrophages to M2 macrophages by secreting interleukin 10 (IL-10), and facilitate tumor cell metastasis and invasion by secreting matrix metalloproteinases.8,9 Several studies have revealed the antitumor potential of COX-2 inhibitors to improve cancer response to immunotherapy at a relatively high dosage.10,11 However, one important caveat is that high dosage of COX-2 inhibitors increase gastrointestinal bleeding and cardiac events, which might cause death.12–14 Therefore, the safety profile of COX-2 inhibitors is a major obstacle and should be rigorously defined prior to broader clinical adoption. Aspirin and its primary metabolite, salicylic acid, have been reported to inhibit COX-2 activation.15–18 However, their side effects at high dosages limit their potential application in cancer therapy.19–21 To overcome these limitations, nanotechnology-based drug delivery systems have been exploited for tumor-selective aspirin delivery.22–25 Inorganic nanomaterials, including gold nanorod, have been used to deliver aspirin through π–π interaction, with the resultant nanoaspirin reducing inflammatory responsiveness of tumor tissues.26 Also, poly(l-lactide-co-glycolide) microspheres27–29 and heparan sulfate30 have been utilized to encapsulate aspirin for cancer treatments. However, these nanoforms of aspirin only showed limited tumor inhibition effects, attributable to unavoidable drug leakage during blood circulation and unresponsive drug release in the tumor tissues.31,32 Further, polymeric prodrug formulation is another way of ensuring tumor-selective drug delivery, in which small-molecule drugs are conjugated covalently to various polymers to enable excellent stability in blood circulation, while the active drugs are released rapidly in response to tumor-specific stimuli by utilizing stimuli-responsive linkers.11,33,34 However, the majority of the current stimuli-responsive polymeric prodrugs were designed to cater only for small-molecule drugs containing amino or hydroxy groups.35,36 Hence, for small-molecule drugs with carboxyl groups like aspirin, there are still not many reports on strategies for preparing their stimuli-responsive polymeric prodrugs. Herein, we synthesized a ROS-responsive aspirin polymeric prodrug (P3C-Asp) through Passerini three-component reaction of aspirin with 4-formylbenzeneboronic acid pinacol ester and 5-isocyanopent-1-yne, followed by azide-alkyne Huisgen cycloaddition "click" reaction of aspirin prodrug with dextran (DEX). Such polymer–aspirin conjugate remained stable in the blood circulation while releasing aspirin or salicylic acid quickly in response to the elevated hydrogen peroxide (H2O2) levels in the tumor cells. We found that P3C-Asp showed much more accumulation in tumor tissues after systemic injection, compared with free aspirin; also, it could suppress murine colorectal tumors (CT26) growth effectively. Further analysis revealed that the P3C-Asp modulated the TME with more CD8+ T cells and M1 macrophages infiltration, as well as less MDSCs infiltration and less PGE2 secretion. Furthermore, the combination of P3C-Asp with aPD-1 eradicated the CT26 tumors (Scheme 1). Thus, our study provides a new method for preparing polymeric prodrugs of small carboxyl molecules and proves that targeting COX-2 pathway with polymer-encased prodrug might present a meaningful fabrication for colorectal cancer therapy. Scheme 1 | Schematic illustration the mechanism of P3C-Asp in combination with aPD-1 in cancer immunotherapy. Due to EPR effect, intravenously injected P3C-Asp accumulates into tumor tissues. Once arrived at the tumor tissue site, aspirin is released in the presence of ROS, inhibits COX-2/PEG2, resulting in reduced proportion of MDSCs, increased proportion of CD8+ T cells and M1 macrophages. When combined with aPD-1, the increased CD8+ T cells leads to killing of all the cancer cells, thereby, eradicating the tumor. Download figure Download PowerPoint Results and Discussion Synthesis and characterization of P3C-Asp Aspirin is a commercial analgesic and anti-inflammatory drug, showing a strong effect in blocking the biosynthesis of cyclooxygenases (COXs), including COX-1 and COX-2.10 Aspirin has been proved to be an effective chemopreventive agent for several solid tumors, including breast cancer, colorectal cancer, and prostate cancer.10 Also, aspirin has been applied to reverse chemoradiotherapy resistance.37,38 However, due to its limitations, such as uncontrollable drug distribution and inadequate drug accumulation in tumor tissues, it is difficult for the drug to perform at therapeutic dosage levels for effective and safe cancer treatments. In order to solve these challenges, we implemented a concept of modular design to construct polymeric aspirin prodrugs (P3C-Asp), as shown in the scheme of Figure 1a. Firstly, the aspirin prodrug, PBAsp (4), was prepared through the Passerini reaction of aspirin (3) with 4-formylbenzeneboronic acid pinacol ester (2) and 5-isocyanopent-1-yne (1), with a phenylboronic acid group exhibiting sensitivity to elevated ROS levels in tumor tissues. Then the PBAsp with reserved acetylene group was conjugated onto DEX modified with azide (5) by azide-alkyne Huisgen cycloaddition "click" reaction. We chose DEX derived from leuconostoc mesenteroides as the polymer carrier due to its excellent hydrophilicity, biocompatibility, and biodegradability.39 The designed P3C-Asp could realize enhanced aspirin accumulation in tumor tissues by the enhanced permeability and retention (EPR) effect, as well as triggered aspirin release from P3C-Asp in response to tumoral-elevated ROS stimulus. We characterized the successful preparation of the PBAsp by proton nuclear magnetic resonance (1H NMR), electrospray ionization mass spectroscopy (ESI-MS), and Fourier-transform infrared (FT-IR) spectra. From the 1H NMR data, the appearance of the methine protons (f) of the p-boronabenzyl ester at δ 6.28 (s, 1H) ppm in deuterated chloroform (CDCl3) demonstrated the successful synthesis of PBAsp (Figure 1b). Also, the PBAsp spectrum from ESI-MS revealed a sharp m/z value at 528.4 due to the mass fragmentation product, [M+Na], of the ester-modified prodrug, which confirmed further, the success of the synthesis ( Supporting Information Figure S3). The FT-IR spectrum of PBAsp proved the existence of the alkynyl group ( Supporting Information Figure S4). For effective conjugation with PBAsp, we first reacted DEX with succinic anhydride to obtain succinic acid–decorated dextran (DEX-SA) ( Supporting Information Figure S5). In the 13C NMR spectrum, the appearance of the characteristic peaks of succinic acid (b = 173.6–174.5, c = 28.9–29.0, d = 28.7–28.0, and e = 176.7–177.0 ppm) demonstrated the successful synthesis of DEX-SA. Further, we reacted DEX-SA with 3-azido-1-propylamine to obtain Dextran-graft-3-azido-1-propylamine (DEX-N3) with the appearance of characteristic 3-azido-1-propylamine peaks (f = 36.7, g = 35.3, and h = 48.6 ppm) in the 13C NMR spectrum, and DEX-N3 peak at λ 2150 of the FT-IR spectrum demonstrated a successful synthesis of DEX-N3 ( Supporting Information Figures S6 and S7). PBAsp was conjugated with DEX-N3 via azide-alkyne Huisgen cycloaddition reaction to obtain the final product, P3C-Asp. The chemical structure of P3C-Asp, dissolved in dimethyl sulfoxide-d6 (DMSO-d6), revealed unique peak patterns in 1H NMR spectrum, as the protons of DEX (a, b, c, d, e, f, g, h, i and j) mainly formed two broad peaks ranged from 3.2 to 5.2 ppm, the methyl protons of pinacol (x) formed a sharp peak at 1.3 ppm, while the DMSO and residual H2O formed two sharp peaks at 2.5 and 3.3 ppm. The methine protons of the p-boronabenzyl ester (u) appeared at 6.0 ppm and the aromatic protons (v) ranged from 7.2 to 8.1 ppm, which confirmed the structure of an intact P3C-Asp. The drug loading content was ∼10 wt %, calculated based on the ratio of the methine proton characteristic peak u to the sum of I, g, h, a, and l (Figure 1c). Figure 1 | Synthesis and characterization of ROS-responsive P3C-Asp. (a) Synthesis routes of P3C-Asp through Passerini reaction and azide-alkyne Huisgen cycloaddition "click" reaction. (b) 1H NMR spectrum of PBAsp in CDCl3. (c) 1H NMR spectrum of P3C-Asp in DMSO-d6. 1H NMR, proton nuclear magnetic resonance; DMSO-d6, deuterated dimethyl sulfoxide. Download figure Download PowerPoint Self-assembly and in vitro drug release of P3C-Asp DEX is a type of classical biodegradable polymer material with excellent water solubility. Due to the hydrophobic effect caused by PBAsp moiety of P3C-Asp, it could self-assemble into nanoparticles readily with a mean hydrodynamic radius of ∼ 40 nm, as evidenced by dynamic light scattering (DLS) and transmission electron microscopy (TEM; Figure 2a and 2b). Such an appropriate size should be conducive to the accumulation of P3C-Asp nanoparticles in tumor tissues. The hypoxic state of tumor tissues leads to elevated ROS level in the tumor cells than that of normal tissues, thereby, accelerating the hydrolysis of P3C-Asp in the tumor cells specifically, with the resultant enhancement of the release of the active drugs, aspirin the deacetylated form, salicylic acid (SA) (Figure 2c). As shown in Figure 2d, at pH 6.8, the release rate of aspirin did not show distinct differences compared with that at pH 7.4. In the presence of H2O2, borate ester groups hydrolyzed easily by oxidation into boric acid, and the newly formed phenolic hydroxyl structure experience intramolecular rearrangement to regenerate the original aspirin or SA. Thus, aspirin released faster at pH 6.8 with 100 μM H2O2 than in the condition of pH 6.8 or pH7.4 without H2O2. In the presence of 100 H2O2 (pH 6.8), > 40% of the total amount of aspirin was released from P3C-Asp nanoparticles in 48 h. Also, ∼ 100% of the released aspirin was converted into SA at the time of detection ( Supporting Information Figure S8). As the concentration of H2O2 reached as high as 10 mM, ∼ 100% of aspirin was released from P3C-Asp, which marked the end of the experiment. Further, it has been reported that aspirin could inhibit the production of ROS by downregulating NADPH oxidase 4 (Nox4) that produces ROS and an inducible nitric oxide synthase (iNOS) that catalyzes the generation of nitric oxide (NO).40,41 We envisaged that the reduction of ROS caused by the intracellular released aspirin should not affect the remaining aspirin release from the prodrug for the following two reasons: (1) Our results for the in vitro drug release assay showed that the 4-formylbenzeneboronic acid linker was fragile to withstand complex physiological conditions even at low H2O2 concentration (100 μM, Figure 2d). (2) The release of aspirin is a gradual process, while tumors are known to generate large amounts of ROS because of their vigorous metabolism and inflammation, which could diminish the effect of aspirin on these free radical related enzymes. In addition, based on the difference of physicochemical properties such as ROS levels between tumors and normal tissues, the ROS-responsive-release behaviors could ensure the tumor-specific release of aspirin from P3C-Asp, which should be conducive for minimizing the side effects of aspirin. Figure 2 | Self-assembly and in vitro drug release of P3C-Asp. (a) Hydrodynamic radius (Rh) of P3C-Asp measured by DLS. (b) Typical morphologies of P3C-Asp measured by TEM. (c) The proposed release mechanism of P3C-Asp in the presence of H2O2. (d) In vitro aspirin release profiles of P3C-Asp in phosphate buffer with Tween 80 (0.2%, w/v) at four conditions: pH 7.4, pH 6.8 with 100 μM H2O2, and pH 6.8 with 10 mM H2O2, n = 3. Download figure Download PowerPoint In vitro cytotoxicity assays CT26 murine colorectal tumor cells were used to evaluate the in vitro cytotoxicity effects of P3C-Asp and free aspirin through cell viability MTT tests. Both aspirin and P3C-Asp showed almost no effect on CT26 tumor cell proliferation even at aspirin concentration as high as 100 μM (Figure 3a and 3b), consistent with previous reports.27,28 The MTT results were reasonable, as the main function of aspirin in cancer treatment is not directly killing the cancer cells, but rather, its effect on inhibition of the inflammatory pathways.42 Figure 3 | In vitro cytotoxicity effects of free aspirin and P3C-Asp on CT26 cells after incubation for 24 (a) or 48 h (b). Download figure Download PowerPoint Biodistribution DEX has been used widely in drug delivery systems due to its excellent hydrophilicity, which aids in improving the drug water solubility.43 We presumed that enhanced permeability of drugs in solid tumors should contribute to readily accumulation, and favor P3C-Asp sustainability, compared with the free aspirin. Thus, we evaluated the biodistribution of P3C-Asp and the free aspirin in BALB/c mice bearing CT26 tumors (n = 3; volume ∼ 300 mm3) to verify the effect of P3C-Asp in the promotion of drug accumulation in tumor tissues. We hydrolyzed the aspirin in the obtained tissues into SA through sodium hydroxide (NaOH) solution, and the content of aspirin was calculated from salicylic acid measured with a combined high-performance liquid chromatography with mass spectrometry (HPLC-MS). The accumulation of aspirin in the P3C-Asp-treated group was about 3.3-fold higher than that of the free-aspirin-treated group at 6 h postinjection and continued to increase to 6.0-fold at 24 h postinjection. In contrast, the accumulation of aspirin in the free-aspirin-treated group showed negligible differences over 24 h, possibly because the peritoneal–plasma barrier was able to slow down the diffusion of drugs from peritoneal to plasma (Figure 4).44 These results demonstrated that P3C-Asp exhibited notable tumor accumulation and extended tumor retention properties. Figure 4 | Measurement of drug concentration in major organs and tumors of CT26 tumor-bearing BALB/c mice after receiving free aspirin and P3C-Asp once. Free aspirin was injected intraperitoneally at a dosage of 10 mg/kg, while P3C-Asp was injected intravenously at the same aspirin dosage. Results were measured with HPLC-MS using aspirin and SA as standards. Data are shown as means ± SD (n = 3). Download figure Download PowerPoint In vivo antitumor and tumor microenvironment (TME) modulation effects have been studied in mice tumor models COX-2-derived prostaglandin E2 (PGE2) is a well-known pro-inflammatory mediator that fosters tumor progression, making them essential bioactive substances that facilitate tumor promotion via remodeling of immune cells with the consequent construction of the tumor-promoting microenvironment.45–47 We considered that the enhanced aspirin accumulation caused by P3C-Asp might be a driving force for the remission of tumor immunosuppressive microenvironment and suppression of tumor growth. To prove this, we carried out in vivo experiments to evaluate the therapeutic potential of P3C-Asp using murine colorectal tumor mice model. CT26 cells (2 × 106) were injected subcutaneously into the abdomen of BALB/c mice. Once the tumor volumes reached ∼ 80 mm3, the mice were divided into three groups (5 mice/group) and received one of the following treatments – Group 1: phosphate-buffered saline (PBS); Group 2: free aspirin; Group 3: P3C-Asp at a dosage of 20 mg/kg based on aspirin, on day 0, 3, and 6 (Figure 5a). We observed that a single treatment with P3C-Asp could slow tumor growth. In contrast, the same dosage of free aspirin treatment in Group 2 showed no discernable impact on tumor growth, which was comparable with the control (PBS) treatment Group 1 (Figure 5b). No noticeable body weight changes were observed in the mice upon receiving different treatments (Figure 5c). Since COX-2 plays a critical role in the immunosuppression of TME, inhibiting the activity of COX-2 should relieve the immunosuppressive TME state effectively. The immune cell population inside the tumor was analyzed on the day after the final treatment (day 7) by ( Supporting Information Figure As the T cells infiltration was observed in the P3C-Asp group, and the M1 macrophages and the M2 macrophages ratio all increased after P3C-Asp demonstrated that aspirin treatment could increase M1 but M2 in an in vitro study when cells were with tumor cells and was a reduced level of infiltration of MDSCs and T cells which are the immunosuppressive cells in TME (Figure and Supporting Information Figure The of MDSCs and and the increase of the ratio of M1 over M2 macrophages the of TME from the protumor type to antitumor type after the P3C-Asp In contrast, the treatment with free aspirin showed a limited on the The content inside the tumor tissues after the final treatment (day 7) were analyzed by assay The results showed that the of and tumor and the 1 type were the in the P3C-Asp-treated group, compared with that of the free-aspirin-treated group and the control Such results were consistent with in immune T cells and M1 macrophages (Figure the content of interleukin 4 a for the of of T cells to cells, in all the groups did not show a significant difference or changes in a content of an anti-inflammatory type and a for the production of was observed after P3C-Asp In contrast, the free group showed a effect on the of the content of PGE2 after the treatment of P3C-Asp, while the treatment of free aspirin showed no effect in the PGE2 content, as compared with the As MDSCs are protumor immune cells in the immunosuppressive TME, and PGE2 is the major secreted by tumor cells to the generation of a in PGE2 should be the for the changes of TME after P3C-Asp treatments. The major including the and were from the three mice groups after the final treatment (day 7) and utilized to evaluate the side effects after receiving different aspirin (5 and the were prepared via and followed by under light to we observed no in the major organs all the treatment groups ( Supporting Information Figure Figure | and TME modulation of P3C-Asp (a) The treatment of the antitumor (b) The tumor growth and tumor suppression rate of mice after treatments with or free aspirin, or n = (c) of mice with different (d) cells in TME after receiving the final treatments on day n = The of in tumor tissues by after receiving the treatments on day and n = no Download figure Download PowerPoint effect of P3C-Asp combined with aPD-1 As shown the treatment of P3C-Asp triggered more CD8+ T cells infiltration and the of immunosuppressive These should be for application as a combined with the immune checkpoint inhibitors To prove we combined or P3C-Asp with aPD-1 2 × murine cancer CT26 cells were injected subcutaneously into the abdomen of BALB/c mice. As the tumor volumes reached ∼ 80 mm3, the mice were divided into groups and group received one of the following Group 1 – Group 2 – aspirin; Group 3 – Group 4 – aspirin Group – and Group 6 – P3C-Asp the groups were injected with a dosage of 20 mg/kg, based on aspirin, on 0, 3, and the groups were injected intraperitoneally with a dosage of 100 aPD-1 on and (Figure As shown in Figure and c, the combined of P3C-Asp with aPD-1 could CT26 In contrast, aPD-1 showed a limited tumor suppression there were no effects between the and aPD-1 and therapeutic effect was exhibited by free aspirin aPD-1 treatment group, compared with aPD-1, the of a tumor-selective delivery in using aspirin for cancer therapy. Further, the results of our body showed negligible changes after all of the various the of the drug (Figure our results demonstrated that P3C-Asp could be a drug delivery in with cancer drug for immunotherapy. Figure 6 | antitumor of P3C-Asp combination with (a) The therapeutic of the in vivo (b) The CT26 tumor growth and tumor suppression rate of mice with free aspirin, aPD-1, free aspirin with aPD-1, P3C-Asp, or P3C-Asp with n = (c) of CT26 tumor-bearing mice with different (d) The CT26 tumor growth for mice after receiving different n = no Download figure Download PowerPoint The and of NMR FT-IR and and the of the used for analysis are all in Supporting We have a new strategy for preparing tumor drug conjugates utilizing Passerini reaction and "click" The carboxyl drug could release from the conjugates (P3C-Asp) and into its original or under the conditions of TME stimuli. single