Dual-aptamer-engineered M1 macrophage with enhanced specific targeting and checkpoint blocking for solid-tumor immunotherapy

Abstract

Chimeric antigen receptor T (CAR-T) cell therapy has faced a series of challenges and has shown very little efficacy in solid tumors to date. Although genetically engineered macrophages have achieved definite therapeutic effect in solid tumors, heterogeneous expression of engineered proteins and the potential for toxicity limit further applications. Herein, we propose a nongenetic and simple macrophage cell engineering strategy through glycan metabolic labeling and click reaction for the treatment of solid tumors. The aptamer-engineered M1 macrophage (ApEn-M1) showed enhanced active targeting ability for tumor cells in vitro and in vivo, resulting in significant cytotoxicity effects. Moreover, ApEn-M1 exhibited superior antitumor efficacy in a breast cancer xenograft mouse model and a lung metastasis mouse model of breast cancer. Interestingly, the ApEn-M1 could reprogram the immunity microenvironment by increasing T cell infiltration and enhancing T cell activity in the tumor region. Additionally, the administration of ApEn-M1 showed no obvious systemic side effects. With glycan metabolic labeling, the macrophages could be efficiently labeled with aptamers on the cell surface via click reaction without genetic alteration or cell damage. Hence, this study serves as a proof of concept for cell-surface anchor engineering and expands the range of nongenetic macrophage cell engineering strategies.

Graphical abstract

Qian and colleagues successfully developed a macrophage cell-surface aptamer-engineering strategy that resulted in a significant cytotoxicity effect and in immunity microenvironment reprogramming for solid-tumor immunotherapy, which serves as a proof of concept for cell-surface anchor engineering and expands the range of nongenetic macrophage cell engineering strategies.

Introduction

Chimeric antigen receptor T (CAR-T) cell therapy is an emerging immunotherapy manner that infuses the patient with CAR-T cells. In particular, two approved CAR-T cell products (Kymriah and Yescarta) have shown unprecedented efficacy in the treatment of certain types of B cell leukemia and lymphoma.^1,2 Despite these encouraging results, CAR-T cell therapy has faced a series of challenges and has shown very little efficacy in solid tumors. The presence of dense stroma and physical barriers, which are difficult to traffic and to penetrate the tumor tissue with CAR-T cells, is one of important reasons for limited therapeutic benefits in solid tumors.^3,4 Additionally, expression of immune-checkpoint ligands and the presence of immunosuppressive cells prevent the activation and expansion of incoming CAR-T cells.^5,6

Previous studies have demonstrated that macrophages are multifunctional and highly plastic cells that are able to penetrate the dense stromal tissue surrounding tumors and actively accumulate in tumor regions.^7,8 In general, macrophages are simplified into two categories: M1, classically activated macrophages, or M2, alternatively activated macrophages. Among others, M1 macrophages exert superior antitumor ability, including the direct killing of tumor cells, tumor antigen presentation, and the promotion of adaptive immune responses.^9,10 Also, M1 macrophages can recruit T helper type 1 (Th1) cells to the tumor site through the secretion of the chemokines CXCL9 and CXCL10.^11,12 Therefore, the distinctive plasticity and versatility of macrophages in combination with their intrinsic capacity to penetrate tumors make them a unique avenue for improving solid-tumor immunotherapy.

In this regard, genetically engineered macrophages have been proposed by CAR methods to specifically recognize cancer cell-surface antigens. For example, Gill et al.¹³ developed CAR macrophages (CAR-Ms) by chimeric adenoviral vector Ad5F35 transduction for five different types of solid-tumor-targeted therapy. Kim et al.¹⁴ established CAR-interferon-γ-encoding plasmid DNA macrophages for the treatment of solid tumors via targeted phagocytosis of cancer cells and antitumor immunomodulation. Vale et al.¹⁵ engineered a family of CAR-Ms with several intracellular signaling domains based on the recognition of defined cell-surface markers to promote specific engulfment and elimination of cancer cells. Undoubtedly, these engineered macrophages generate great antitumor activity in solid tumors. However, this genetic-engineering strategy may face some potential problems and risks, including potential target organs for toxicity, heterogeneous expression of engineered proteins, and transgene insertional mutagenesis.^16,17 These issues inevitably reduce the therapy efficacy of this genetic manner and hinder its further application. In addition, the production of genetically engineered cells is laborious and costly, involving most-complicated procedures. Therefore, there is an urgent need to develop simpler and safer technology to address the challenges associated with macrophage therapies.

Herein, inspired by favorable features and the wide applicability of metabolic glycan labeling, we proposed an aptamer-engineered macrophage strategy for enhanced immunotherapy in solid tumors. Aptamers are short single-stranded DNA (ssDNA) or RNA with distinct secondary and tertiary structures, which could bind to target molecules with high specificity and affinity.¹⁸ In this study, a mouse mononuclear macrophage leukemia cell line (RAW264.7) was first incubated with Ac₄ManNAz and lipopolysaccharide (LPS) to generate azido sugars labeling M1 macrophages (Figure 1). Ac₄ManNAz could incorporate into the glycans of the cell-surface protein via cellular sugar-metabolic synthesis without disturbing the cells’ physiological functions.¹⁹ Next, dibenzocyclooctyne-modified AS1411 aptamers (DBCO-AS1411) and dibenzocyclooctyne-modified PD-L1 aptamers (DBCO-PD-L1) were anchored on the cell’s surface through the simple and efficient azido-alkenyl group’s click-chemistry reaction process. AS1411 aptamers specifically recognized nucleolin overexpressing on the surface of many cancer cells.²⁰ Likewise, PD-L1 aptamers could bind to PD-L1 in the plasma membrane of multiple solid cancer types and simultaneously serve as immune-checkpoint inhibitors.²¹ Under surface-anchored aptamer guidance, the aptamer-engineered M1 macrophages (ApEn-M1) could specifically bind to cancer cells and promote tumor eradication. Taken together, this rational combination of metabolic glycan labeling and macrophage therapy provided an alternative strategy for boosting solid-tumor therapy.

Figure 1.

System design and immunotherapeutic effects of ApEn-M1

(A) Schematic illustration of preparation of ApEn-M1 via metabolic glycan biosynthesis and click-chemistry reaction. (B) Schematic illustration of ApEn-M1 for enhanced immunotherapy in tumor-bearing mice.

Results

Generation of ApEn-M1

To functionalize M1 macrophage cells for cancer therapy, AS1411 and PD-L1 aptamers were anchored to the cell surface via metabolic glycan biosynthesis and click-chemistry reaction. Confocal-microscopy results revealed obvious green (FAM-AS1411) and red (Cy3-PD-L1) fluorescence signals on the RAW264.7 cell’s surface (Figure 2A). In addition, a random control sequence (FAM-rAS1411, Cy3-rPD-L1) group also showed intense fluorescence signals. The z stack image of ApEn-M1 showed AS1411 and PD-L1 aptamers located in the cell membrane (Figure S1). The phenomenon demonstrated that the aptamers were anchored to the cell surface by covalent binding of a bioorthogonal reaction. Similarly, flow-cytometry results also indicated that the aptamers were successfully attached to the cell surface (Figures 2B and 2C). Further, the security of the metabolic glycan label was investigated by CCK-8 assay. As shown in Figure S2, no obvious reduction in RAW264.7 cell viability was observed when incubated with Ac₄ManNAz at 24 and 48 h.

Figure 2.

Generation and verification of ApEn-M1

(A) Confocal-microscopy images of RAW264.7 cells after incubation with different aptamers and Ac₄ManNAz (scale bar: 20 μm, 40× magnification). (B and C) Flow-cytometry analysis of RAW264.7 cells after incubation with different aptamers and Ac₄ManNAz. (D) Confocal-microscopy images of 4T1 cells after incubation with different aptamers (scale bar: 20 μm, 40× magnification). (E and F) Flow-cytometry analysis of 4T1 cells after incubation with different aptamers.

To confirm the aptamer stability in serum, RAW264.7 cells decorated with FAM-AS1411 aptamer were incubated with fetal bovine serum for different amounts of time. Meanwhile, the fluorescent signals of ApEn-M1 surface-decorated aptamers were monitored. As shown in Figures S3 and S4, flow-cytometry and confocal-microscopy results revealed that ApEn-M1 surface-anchored aptamer signals remained relatively stable for up to 24 h.

Next, we investigated the targeting ability of free AS1411 and PD-L1 aptamers for 4T1 cells. As shown in Figure 2D, obvious fluorescence signals were observed from 4T1 cells. In contrast, the rAS1411 and rPD-L1 aptamer groups showed weak fluorescence signals. Also, flow-cytometry results reflected a substantially higher number of positive cells in the AS1411 and PD-L1 aptamer groups compared with the control and rAS1411 and rPD-L1 aptamer groups (Figures 2E and 2F). Additionally, we investigated the aptamer’s specificity using another positive cell line (MDA-MB-231) and a control cell line (HEK-293T). As shown in Figure S5, flow-cytometry and confocal-microscopy results revealed that AS1411 and PD-L1 aptamers possess high binding affinity and good specificity.

PD-L1 aptamer function investigation

Further, PD-L1 aptamer functions were also explored in the study, and a schematic diagram depicts this detailed process (Figure 3A). We first overexpressed PD-1 via plasmid transfection in the PD-1-negative cell line 4T1; the result was confirmed by western blot (Figure 3B). Subsequently, PD-L1 aptamers and hFc-tagged recombinant PD-L1 protein were incubated with PD-1-overexpression 4T1 cells. After washing, the cells were stained with phycoerythrin (PE) anti-human immunoglobulin G (IgG)-Fc antibody and analyzed by flow cytometry. As shown in Figures 3C and 3D, the mean fluorescence intensity significantly decreased in the presence of PD-L1 aptamers. These results indicate that PD-L1 aptamers could interfere in the PD-1/PD-L1 interaction.

Figure 3.

Functional study of PD-L1 aptamer

(A) Schematic illustration of flow-cytometry analysis of PD-L1 aptamer interfering PD-L1 interaction with PD-1 on 4T1 cell surface (Fc-PD-L1, recombinant human Fc-tagged mouse PD-L1 protein). (B) Western blot of PD-1 plasmid transfection. (C and D) Flow-cytometry analysis of PD-L1 aptamer interfering with the PD-1/PD-L1 interaction on 4T1 cell surface. The error bars are standard deviations of three repetitive measurements. For all panels, ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001.

Antitumor effects of ApEn-M1 in vitro

Previous studies have reported that M1 macrophages actively phagocytose tumor cells and release tumor necrosis factor α (TNF- α), reactive oxygen species (ROS), and nitric oxide (NO) for the direct killing of tumor cells.^22,23 We first investigated the phagocytic ability of ApEn-M1 for target cells. Confocal-microscopy results showed that most 4T1 cells were phagocytized by ApEn-M1 in comparison with M1 cells, which attributed to cell-surface aptamers facilitating and prolonging interactions between ApEn-M1 and tumor cells (Figure 4A). To further explore this therapeutic effect, we prepared rAS1411 and rPD-L1 aptamer-engineered M1 cells (M1), AS1411 and rPD-L1 aptamer-engineered M1 cells (A-M1), PD-L1 and rAS1411 aptamer-engineered M1 cells (P-M1), and AS1411 and PD-L1 aptamer-engineered M1 cells (A-P-M1 or ApEn-M1, respectively). In particular, the A-P-M1 group represented a higher phagocytic proportion compared with the A-M1 or P-M1 group, indicating an enhanced active targeting ability. Consistently, flow-cytometry results showed that the percentage of dual-positive cells was significantly increased in the A-P-M1 group, suggesting significantly higher phagocytosis of cancer cells than other groups (Figures 4B and 4C). Moreover, the viability of target cells gradually decreased along with the ratio of 4T1/ApEn-M1 cell reduction (Figure S6). Also, the A-M1 cells showed little cytotoxicity for cell-surface nucleolin-negative HEK-293T cells, indicating the importance of the selective targeting of tumor cells (Figure S7). We further confirmed the targeted cytotoxicity of ApEn-M1 by co-incubating 4T1 cells and ApEn-M1. As shown in Figure 4D, the viability of target cells significantly reduced with ApEn-M1 treatment. Calcein-AM staining also confirmed that 4T1 cells significantly reduced the proliferative activity in the A-P-M1 group (Figure S8). Simultaneously, flow-cytometry results revealed that ApEn-M1 treatment induced a significant increase in the percentage of apoptosis cells, especially the A-P-M1 group, which showed the highest apoptosis-cell rate (Figures 4E and 4F). Taken together, these results demonstrate the therapeutic potential of ApEn-M1 to actively target tumor cells.

Figure 4.

Active targeting therapeutic effect of ApEn-M1

(A) Representative confocal images for phagocytosis of ApEn-M1 (scale bar: 20 μm, 20× magnification). (B) Flow-cytometry results for phagocytosis of ApEn-M1. (C) Percentage of phagocytic in different treatment groups. (D) Cell vitality was detected by CCK-8 assay at 24 h. (E) Cell apoptosis was analyzed by flow cytometry at 24 h. (F) Percentage of apoptosis cells in different treatment groups. The error bars are standard deviations of three repetitive measurements. For all panels, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001.

Biodistribution of ApEn-M1 in vivo

Encouraged by these in vitro results, we further evaluated the ApEn-M1 potential therapeutical effect in a breast cancer xenograft mouse model. First, different Cy7-labeled aptamer-engineered RAW264.7 cells were intravenously injected into BALB/c mice through the tail vein to evaluate the active targeting ability and biodistribution of ApEn-M1. After 12 h, an in vivo bioluminescence imaging system displayed a significantly fluorescence signal in the tumor area (Figures 5A and 5B). Nonetheless, the M1 group had almost no fluorescence signal due to the lack of active targeting capability. Notably, the dual-aptamer-modified M1 (A-P-M1) exhibited higher fluorescence intensity than that of the single aptamer-modified M1 (A-M1 or P-M1), indicating an enhanced active targeting ability for A-P-M1. Furthermore, the fluorescence intensity of tumor tissues could remain stable for at least 48 h post-injection of A-P-M1 (Figures 5C and 5D).

Figure 5.

In vivo fluorescence images of ApEn-M1 (n = 5)

(A) Representative fluorescence images of the tumor-bearing mice from different groups. (B) Relative fluorescence intensity of tumors in different groups. (C) Representative fluorescence images of the tumor-bearing mice at different time points. (D) Relative fluorescence intensity of tumors in different time points. (E) Representative fluorescence images of main organs and tumors. (F) Relative percentages of fluorescence intensity in main organs and tumors. For all panels, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001.

We next sought to evaluate the biodistribution of ApEn-M1 by main ex vivo organ imaging at 12 h post-injection. For the M1 group, the fluorescence signal was mainly accumulated in liver and lung (Figures 5E and 5F). As expected, the proportion of fluorescence signal significantly increased in tumor tissues for the ApEn-M1 group. Similarly, the proportion of fluorescence signal in tumor tissues was higher in the A-P-M1 compared with the A-M1 or P-M1 groups.

Antitumor effects of ApEn-M1 in vivo

On this basis, the tumor-bearing mice received intravenous injections of the ApEn-M1 at different times, and the treatment process is schematically illustrated in Figure 6A. As expected, we found that ApEn-M1 could induce significant regression in tumor volume and improve survival rate compared with control and M1 cells group (Figures 6B–6D). Simultaneously, the A-P-M1 group showed smaller tumor sizes and higher survival rates than the A-M1 or P-M1 group. To further confirm therapeutic efficacy, hematoxylin and eosin (H&E), TUNEL, and immunohistochemical stainings for antigen Ki-67 tumor were carried out for each group of tumor-tissue slices. As shown in Figure 6E, the most necrotic area with obvious nucleus shrinkage was observed in the images of tumor-tissue slices with H&E staining from the A-P-M1 group. Meanwhile, TUNEL staining showed that the number of apoptosis cells was the largest in the A-P-M1 group. Similarly, the A-P-M1 group showed a smaller positive rate than the other groups for Ki-67 assays, indicating the weak proliferation levels of tumor cells.

Figure 6.

ApEn-M1 treatment effectively inhibited tumor progression in vivo (n = 5)

(A) Illustration of the treatment process. (B) Photographs of the tumor tissue from each group (n = 5). (C and D) Tumor-growth profiles and survival percentages of mice receiving different treatments (n = 5). (E) H&E, TUNEL, and Ki-67 staining of tumor-tissue slices at the time point 24 days of treatment (scale bar: 25 μm). For all panels, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001.

In view of the above results, we assessed the effects of ApEn-M1 treatments on tumor immune microenvironment reprogramming by immunohistochemical assay and flow cytometry. On the basis of immunohistochemical-assay results (Figure S9), the transferred ApEn-M1 group showed significantly enhanced M1 macrophage (inducible nitric oxide synthase [iNOS]⁺), CD8⁺, and CD4⁺ T cell infiltration accompanied by reduced M2 macrophage (Arg-1⁺) and regulatory T (Treg) cell (Foxp-3⁺) infiltration in tumor tissues. In general, inducible iNOS expression is an important marker for M1 macrophages, arginase-1 (Arg-1) expression is an important marker for M2 macrophages, and Foxp-3 expression is an important marker for Treg cells. Also, M1 macrophages activate immune responses and tumoricidal activity.²⁴ Conversely, M2 macrophages suppress adaptive immunity and facilitate tumor growth and progression.²⁵ Moreover, flow-cytometry results displayed that the percentages of both CD3⁺ CD8⁺ and CD3⁺ CD4⁺ T cells were much higher in the A-P-M1 group than in the other groups (Figure 7). Notably, the percentage of granzyme B⁺ CD8⁺ T cells and Ki-67⁺ CD8⁺ T cells also obviously increased in tumor tissue of mice treated with A-P-M1. Taken together, the ApEn-M1 therapy strategy improved active targeting ability and anti-tumor efficiency, increased tumor-infiltrating T cells, and enhanced cytotoxicity T cell activities.

Figure 7.

Representative flow-cytometric plots showing CD4⁺ T cell and CD8⁺ T cells ratios and function in tumor tissue at the time point 24 days of treatment

For all panels, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001.

To evaluate the therapeutic safety, the vital organs of mice were harvested and stained by H&E. As shown in Figure S10, no obvious histopathologic changes were observed in each group, confirming the biocompatibility of the ApEn-M1 treatment.

Antitumor effects of ApEn-M1 in lung metastasis tumors

Encouraged by the promising antitumor effect of ApEn-M1, we extended its application into the treatment of a lung metastasis mouse model of breast cancer. As shown in Figure 8A, the mice were sacrificed, and the lungs were collected after 14 days of intravenous tumor-cell injection. M1 cell treatment showed only mild antimetastatic effects compared with the control group (Figure 8B). In contrast, ApEn-M1 therapy significantly inhibited lung-metastasis progression, including reduced metastatic nodules and lung weight (Figures 8C and 8D). Furthermore, H&E staining showed a distinctly reduced tumor area with ApEn-M1 treatment, suggesting the effective inhibition of lung metastasis (Figure 8E). Notably, the A-P-M1 group showed fewer metastatic nodules and less lung weight and tumor area than the A-M1 or P-M1 group, which can be attributed to the enhanced interaction with cancer cells in the circulatory system.

Figure 8.

ApEn-M1 treatment effectively inhibited tumor lung metastasis progression (n = 5)

(A) Illustration of the treatment process. (B) Images of the lung metastatic lesions (n = 5). (C and D) The number of metastatic nodules and the lung weight for each group (n = 5). (E) H&E staining of the lung metastatic lesions. For all panels, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001.

Discussion

Macrophages are highly plastic cells and regulators of the innate immune system, capable of phagocytosis, cellular cytotoxicity, secretion of pro-inflammatory factors, and antigen presentation to T cells.^24,26, 27, 28, 29 These functions are critical in the response to bacterial and viral pathogens and have the potential to participate in antitumor immunity.30, 31, 32 In the present study, we have developed a macrophage cell-surface aptamer-engineering strategy to boost solid-tumor immunotherapy. ApEn-M1 exerted predominant targeting ability for cancer cells under surface-anchored aptamer guidance and triggered cancer cell apoptosis. Moreover, this rational combination of aptamer-engineering and M1 cell therapy yielded improved therapy efficacy in a breast cancer xenograft mouse model and a lung metastasis mouse model of breast cancer. Notably, ApEn-M1 showed the potential to further reprogram the tumor microenvironment via increased tumor-infiltrating T cells and enhanced cytotoxicity T cell activities.

As we all know, macrophage cells are resistant to genetic-engineering approaches, thus limiting their therapeutic effect for clinical application.^7,33,34 Bioorthogonal chemical strategies have been recently successfully applied for labeling and imaging in living cells.35, 36, 37, 38 In this study, we developed a unique approach to render macrophages cell target specificity by simply anchoring aptamers on the cell surface without genetic alteration or cell damage. With glycan metabolic labeling, the macrophages could be efficiently labeled with aptamers via click reaction.

However, the ApEn-M1 cells inevitably encounter unfavorable conditions in the tumor microenvironment, such as a hypoxic and acidic environments, expression of immune-checkpoint ligands, and an abundance of immunosuppressive cells.39, 40, 41, 42 These factors facilitate ApEn-M1 cells toward M2 phenotype polarization. Given these problems, we need to endow the ApEn-M1 cells with a robust pro-inflammatory polarization signature with the potential to further polarize the surrounding microenvironment, including metal ion, small-molecule drugs, and so on. In this study, we found that the ApEn-M1 cell treatment increases tumor-infiltrating T cells and enhanced cytotoxicity T cell activities. Hence, we will improve the ApEn-M1 strategy and study the cross-presentation and co-stimulation of ApEn-M1 and T cells in future work.

Materials and methods

Materials

All aptamer sequences were synthesized and purified by Sangon Biotechnology (Shanghai, China). Detailed information on aptamers sequences are shown in Table S1. Antibodies for iNOS, Arg-1, CD4, CD8, Ki-67, and Foxp-3 were obtained from Servicebio (Wuhan, China). The DAB TUNEL Cell Apoptosis Detection Kit was purchased from Servicebio (Wuhan, China). N-Azidoacetylmannosamine-tetraacylated (Ac₄ManNAz) was purchased from Thermo Fisher Scientific (Cambridge, MA, USA). Cell Counting Kit (CCK-8) was obtained from MedChemExpress (Monmouth Junction, NH, USA). Calcein-AM, calcein Red-AM, anti-mouse CD16/32 antibody, PE anti-human IgG-Fc antibody, PerCP/Cyanine5.5 anti-mouse PD-L1 antibody, fluorescein isothiocyanate (FITC)-CD4, FITC-CD8, APC-CD3, APC-Ki-67, and APC-granzyme B were purchased from BioLegend (San Diego, CA, USA). Annexin V-FITC/PI Apoptosis Detection Kit was purchased from Beyotime (Haimen, China). PD-1 plasmid was prepared by GeneChem (Shanghai, China). Mouse recombinant PD-L1 protein with hFc tag was purchased from Sino Biological (Beijing, China). Protease inhibitors were purchased from TargetMol (Boston, MA, USA). Ultrapure LPS was purchased from InvivoGen (San Diego, CA, USA). All oligonucleotides were dissolved in sterile water and stored at −20°C.

Cell culture

Mouse breast cancer cell line 4T1, mouse macrophage cell line RAW264.7, human breast cancer cell line MDA-MB-231, and human embryonic kidney cell line HEK 293T were obtained from the American Type Culture Collection (ATCC; Rockville, MD, USA) and were cultured in RPMI 1640 medium (Gibco, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (Gibco) and 1% penicillin/streptomycin. All cell lines were incubated in an atmosphere of 5% CO₂ at 37°C according to ATCC guidelines.

Preparation of ApEn-M1

RAW264.7 cells were treated with LPS (100 ng/mL) and Ac₄ManNAz (50 μM) for 24 h and washed with PBS 3 times. Then, the cells were incubated with DBCO-decorated FAM-AS1411 and Cy3-PD-L1 aptamers in fresh RPMI 1640 medium at 37°C for 1 h. The obtained ApEn-M1 was directly used for in vitro and in vivo studies.

To confirm that the aptamer successfully anchored to the cell surface, the cells were fixed with 4% formaldehyde (Solarbio, Beijing, China) for 15 min at room temperature, and cell nuclei were treated with DAPI for 10 min. Finally, the cells were analyzed by confocal microscopy (Leica TCS SP8, Germany) and flow cytometry (Beckmann Coulter CytoFLEX, USA) to confirm that the aptamer successfully bound with the cells.

Targeting ability of aptamer

To investigate the active targeting ability of the aptamer, 4T1, MDA-MB-231, and HEK 293T cells were grown on a glass-bottom dish (35 mm with 15 mm bottom well) at a density of 5 × 10⁵ cells/mL and were incubated at 37°C for 24 h. After being washed with PBS buffer three times, the cells were incubated with FAM-AS1411 and Cy3-PD-L1 aptamers in fresh RPMI 1640 medium at 37°C for 1 h. Then, the cells were washed with PBS buffer three times and imaged by confocal microscopy. In addition, cellular fluorescence signals were also quantified by flow-cytometry analysis.

Stability analysis of aptamer

To test the aptamer stability, RAW264.7 cells decorated with FAM-AS1411 aptamer were incubated with fetal bovine serum at 37°C for different time periods (0, 2, 4, 6, 8, 12, 16, and 24 h). Then, the fluorescence intensity was analyzed by confocal microscopy and flow cytometry.

Cellular cytotoxicity assays

RAW264.7 cells were seeded in a 96-well plate with 5 × 10³ cells per well and were incubated with 50 μM Ac₄ManNAz for 24 or 48 h, respectively. Then, the medium was replaced by fresh culture medium, CCK-8 solution was added, and it incubated for 1 h. The absorbance at 450 nm was measured using a Model 680 microplate reader (Bio-Rad, Hercules, CA, USA).

Western blot

Whole protein from 4T1 cells was harvested with radioimmunoprecipitation assay (RIPA) buffer containing proteinase and phosphatase-inhibitor cocktail. The protein concentration was measured by BCA assay. Protein samples were separated on SDS-polyacrylamide gels and transferred onto polyvinylidene difluoride (PVDF) membranes. The membranes were blocked with 5% non-fat dry milk in Tris-buffered saline with 0.05% Tween 20 (TBST) for 1 h followed by incubation at 4°C overnight with the primary antibodies. HRP anti-rabbit IgG was used as the secondary antibody. The signals were analyzed using an enhanced chemiluminescence detection system (Bio-Rad).

Plasmid transfection and PD-L1 aptamer function assay

4T1 cells were transfected with DNA plasmid expressing mouse PD-1 with Lipofectamine 2000 (Invitrogen, Grand Island, NY, USA). Then, the cells were washed with PBS and incubated with 100 μL recombinant Fc-tagged mouse PD-L1 protein (100 μg/mL) and PD-L1 aptamer (2 μM) at 37°C for 30 min. After washing three times with PBS, the cells were stained with PE-tagged anti-human IgG-Fc antibody for 30 min. Finally, the cells were washed with PBS buffer three times and analyzed by flow cytometry.

Similarly, 4T1 cells also were incubated with PerCP/Cyanine5.5-tagged PD-L1 antibody and PD-L1 aptamer for 30 min. Subsequently, the cells were washed with PBS buffer three times and analyzed by flow cytometry.

Transwell assays

ApEn-M1 and 4T1 cells were co-cultured using the Transwell system (Millipore, Bedford, MA, USA) with a 3 μm pore size. Then, ApEn-M1 (3 × 10⁵ cells) was added to the upper chamber, and 4T1 cells (1 × 10⁵ cells) were added to the lower chamber. After incubation for 24 h, the cell viability of 4T1 cells in the lower chambers was determined by a CCK-8 assay.

Phagocytosis assay

3 × 10⁵ ApEn-M1 and 1 × 10⁵ 4T1 cells were, respectively, stained with calcein-AM and calcein Red-AM. In brief, ApEn-M1 (3 × 10⁵ cells) and 4T1 cells (1 × 10⁵ cells) were incubated with different dyes for 20 min at 37°C and kept protected from light. Subsequently, the cells were washed with PBS buffer three times and mixed in fresh culture medium. After incubation for 24 h, the mixed cells were analyzed by confocal microscopy and flow cytometry.

Apoptosis assay

ApEn-M1 (3 × 10⁵ cells) and 4T1 cells (1 × 10⁵ cells) were mixed and incubated for 24 h. Then, 4T1 cells were harvested with EDTA-free trypsin and washed with PBS three times. Subsequently, the cells were resuspended in 500 μL PBS and stained with 5 μL Annexin V- FITC and 5 μL propidium iodide (PI) for 15 min in the dark at room temperature. Finally, the apoptosis of 4T1 cells was evaluated by flow cytometer.

Animal studies

Female BALB/c mice (5–6 weeks) were purchased from the Experimental Animal Center of Chongqing Medical University and used following protocols approved by the Ethics Committee of Chongqing Medical University. A total of 1 × 10⁶ 4T1 cells in 100 μL PBS were subcutaneously injected into the right flank of each mouse.

For in vivo imaging, the mice were intravenously injected in the tail vein with M1, A-M1, P-M1, or A-P-M1 (5 × 10⁶ cells) to evaluate the active targeting ability of different groups (n = 5). M1 was labeled with rAS1411-Cy7 and rPD-L1-Cy7, A-M1 was labeled with AS1411-Cy7 and rPD-L1-Cy7, P-M1 was labeled with rAS1411-Cy7 and PD-L1-Cy7, and A-P-M1 was labeled with AS1411-Cy7 and PD-L1-Cy7. The fluorescence images were captured by an in vivo imaging system (LB983 NightOWLII, Berthold Technologies) for 12 h. To observe ApEn-M1 accumulation on the tumor site, the mice were intravenously injected in the tail vein with A-P-M1 (5 × 10⁶ cells), and fluorescence intensity was monitored at 3, 6, 12, 24, and 48 h. To evaluate the biodistribution, the mice were intravenously injected in the tail vein with M1, A-M1, P-M1, or A-P-M1 (5 × 10⁶ cells). At 12 h, mice were euthanized, and fluorescence images of the main organs and tumors were performed (n = 5).

To investigate the antitumor effect of ApEn-M1, a total of 1 × 10⁶ 4T1 cells in 100 μL PBS were subcutaneously injected into the right flank of each mouse. After 2 weeks, the tumor-bearing mice were randomly divided into 5 groups (5 mice per group) and intravenously injected in the tail vein with PBS, M1, A-M1, P-M1, or A-P-M1 (5 × 10⁶ cells), respectively. Altogether, the tumor-bearing mice were treated 3 times, and the treatment time points are shown in Figure 5A. Tumors were measured once every 3 days with digital calipers, and tumor volumes were calculated by the following formula: volume = length × (width)²/2.

To evaluate the potential toxicity and side effects, the mice were sacrificed, and the main organs were removed to be stained with H&E. The images were observed by microscopy (Nikon ECLIPSE Ti-s, Japan).

For the lung metastasis model, 1 × 10⁶ cells suspended in 100 μL PBS were injected intravenously in the tail vein into BALB/c mice. Two days later, the tumor-bearing mice were randomly divided into 5 groups (5 mice per group) and intravenously injected in the tail vein with PBS, M1, A-M1, P-M1, or A-P-M1. Altogether, the mice were treated 3 times, and the treatment time points were the 3^rd, 5^th, and 7th days. After treatment for 2 weeks, the mice were sacrificed, and the lungs were weighed and fixed with 4% formaldehyde. Then, the numbers of metastatic nodules in the lungs were examined.

Immunohistochemistry

After the 4T1-bearing mice sacrificed, tumors were excised and fixed in 4% formaldehyde, embedded in paraffin, and then sliced into 4 μm sections. After deparaffinating and immunostaining with iNOS, Arg-1, CD4, CD8, Ki-67, and Foxp-3 primary antibodies and rabbit anti-mouse IgG secondary antibodies, expression levels of those antigens were then detected by HRP-conjugated DAB.

Flow-cytometry analysis of immune cells

At the time point 24 days of treatment in each group, the tumors were collected and digested to obtain single-cell suspensions. Then, the cells were stained with the FITC-CD4, FITC-CD8, APC-CD3, APC-Ki-67, and APC-granzyme B antibodies for 1 h. Finally, single-cell suspensions were evaluated by flow cytometer.

Statistical analysis

Student’s t test or one-way analysis of variance (ANOVA) was used to assess the differences between treated and control groups with GraphPad Prism 8.0.1. All quantitative results are presented as mean ± SD (standard deviation) with at least three independent repetitive experiments.