Preparation and characterization of M@SINPsFig. 1

(A) Construction of M@SINPs. (B) TEM images of SNPs, SINPs and M@SINPs. (C) Stability assay of SNPs, SINPs and M@SINPs in 10% FBS over time (n = 3). (D) FCM analysis of the cellular internalization of free IR820, SINPs and M@SINPs in RAW264.7 cells (n = 3). (E) In vivo imaging of free IR820, SINPs and M@SINPs after i.v. injection at indicated time points. (F) Ex vivo fluorescence analysis in the tumors tissues collected at 12 h post-treatment (n = 3). (**p < 0.01, ***p < 0.001)

The preparation of M@SINPs briefly contained two steps as illustrated in Fig. 1A. HSA and SHP099 were firstly assembled into SHP099-loaded HSA nanoparticles (SNPs) using a desolvation method [37]. After that, the obtained SNPs were sequentially IR820 loaded (SINPs) and mannose modified to prepare the final M@SINPs. Representative transmission electron microscope (TEM) images showed the spherical morphology of the prepared nanoparticles with a uniform distribution (Fig. 1B). Dynamic light scattering determination revealed that SNPs, SINPs and M@SINPs all had an average diameter around 140 nm with a polydispersity index (PDI) of ~ 0.1 (Table S1). Meanwhile, the zeta potential of SINPs changed after mannose modification, which further indicated the successful conjugation of mannose (Figure S1). The loading of SHP099 and IR820 for M@SINPs was 43.2 ± 0.3 and 24.7 ± 0.1 µg mg− 1 with the loading efficiency of 86.4% and 98.9%, respectively, by high-performance liquid chromatography (HPLC) and UV detection (Table S1). The obtained M@SINPs remained stable in the size for 7 days upon incubation with 10% fetal bovine serum (FBS) (Fig. 1C). Then, we analyzed the drug release kinetics profiles of these nanoparticles, and all nanoparticles displayed a sustained release of SHP099 and IR820, respectively (Figure S2).

Next, we evaluated the ability of mannose modification to increase the intracellular internalization of M@SINPs in M2 polarized RAW264.7 macrophages (M2-RAW264.7). For the study of the intracellular uptake of the nanoparticles by macrophages, free IR820, SINPs and M@SINPs were incubated with M2-RAW264.7 for 4 h, and the intracellular internalization of the nanoparticles was evaluated quantitatively and qualitatively. Figure 1D showed that M@SINPs were more efficiently internalized by M2-RAW264.7 (about 1.3 times as much as the SINPs). Confocal laser scanning microscopy (CLSM) observation further confirmed this result. As shown in Figure S3, compared with free IR820 and SINPs, the cells incubated with M@SINPs showed more obvious fluorescence signals.

Furthermore, we studied the in vivo tumor accumulation of M@SINPs in CT26-bearing mice. As compared with free IR820 and SINPs, M@SINPs exhibited a prolonged tumor retention time and the stronger fluorescence signal of M@SINPs still could be observed even at 120 h after injection (Fig. 1E). The fluorescence intensity of IR820 at tumor site was 2.20- and 1.44-fold higher in the M@SINPs group than free IR820 and SINPs at 12 h after the injection (Fig. 1F), respectively, suggesting that M@SINPs has relatively better tumor targeting properties. In addition, tumor tissues were also harvested for assessment whether TAMs can be targeted by i.v. M@SINPs administration using flow cytometry (FCM). Figure S4 shows that the M@SINPs can enhance intracelluar internalization of the nanoaprticles in TAMs in tumor site. Overall, the results demonstrated relatively good tumor-targeting properties of M@SINPs in vitro and in vivo, respectively.

M@SINPs facilitated macrophages phagocytosis and M2-to-M1 repolarizationFig. 2

M@SINPs for increasing macrophages phagocytosis. (A) Schematic description of cell phagocytosis experiment. (B-D) FCM analysis of the phagocytosis of CT26 cells by RAW264.7 after treatment with free SHP099, SNPs or M@SINPs (n = 3). (E) Fluorescence images showing the phagocytosis of CTG-labeled CT26 cells (green) by Rhodaming B-labeled RAW264.7 (red). Scale bar, 50 μm. (***p < 0.001)

SHP099, as a SHP2 allosteric inhibitor, has shown promising results in facilitating the engulfment of tumor cells by macrophages in the previous works [23, 24]. To determine whether SHP099-loaded HSA nanoparticles could promote the phagocytosis by macrophages, CT26 cells were marked with cell tracker green (CTG, green) and then co-cultured with RAW264.7 in vitro, which had been pre-treated with the different SHP099 formulations (Fig. 2A). As detected by FCM analysis, compared with the control and free SHP099 group, more cancer cells were phagocytosed by RAW264.7 macrophages pre-incubated with SNPs and M@SNPs (Fig. 2B-D), thus indicating that SHP099-loaded HSA nanoaprticles can promote the phagocytosis of tumor cells by macrophages. Furthermore, confocal imaging further revealed that SNPs and M@SNPs significantly facilitated the phagocytosis of CT26 cells by RAW264.7 macrophages (Fig. 2E).

Fig. 3

M@SINPs for reprogramming M2-to-M1 repolarization in vitro. (A to B) FCM analysis of the expression of M1 markers (CD86+) and M2 markers (CD206+) in M2-RAW264.7 with the treatments as indicated (n = 3). (C-F) Cytokines levels of TNF-α, IL-6, TGF-β and IL-10 in the culture medium of M2-RAW264.7 with the treatments as indicated (n = 3). (*p < 0.05, **p < 0.01, ***p < 0.001)

Intracellular ROS generation can affect the macrophage polarization and activation via the ROS/NF-kB signal pathway [13, 14]. To explore whether the prepared nanoparticles could repolarize M2-TAMs to M1 phenotype, we first evaluated the possible toxicity of the prepared nanoparticles and intracelluar ROS generation in RAW264.7 cells under laser irradiation. For cytotoxicity assay, RAW264.7 cells were incubated with free IR820, free SHP099, SNPs, SINPs and M@SINPs, and then irradiated for different time under laser irradiation (808 nm, 0.5 W/cm2). As indicated in Figure S5, all the formulations remained viable more than 80% under laser irradiation (0.5 W/cm2, 90s). Furthermore, we investigated the intracellular ROS generation in RAW264.7 cells using 2’,7’-dichlorodihydrofluorescein diacetate (DCFH-DA) probe. As shown in Figure S6, SNPs group showed no obvious green fluorescence. SINPs group and M@SINPs group showed dramatically increased green fluorescence relative to free IR820 group, thus confirming the effective intracelluar ROS photo-generation in RAW264.7 cells treated with the prepared IR820-loaded nanoparticles.

Upon confirmation of the intracelluar ROS photo-generation, we next evaluated the role played by the prepared nanoparticles on influencing macrophages polarity. Macrophages can express specific phenotypic biomarkers. For example, CD86, IL-6, iNOS and TNF-a were considered as M1-phenotypic biomarkers, whereas CD206, TGF-b, and IL-10 were considered as M2-phenotypic biomarkers. In this study, SINPs and M@SINPs treatments, upon 808 nm laser irradiation (0.5 W/cm2, 90 s), showed the increased expression levels of CD86, TNF-α and IL-6 and the reduced expression levels of CD206, TGF-β and IL-10 compared with SINPs and M@SINPs groups without laser irradiation, respectively (Fig. 3). These comparisons suggested that the intracellular ROS photogeneration could facilitate the M2-to-M1 repolarization of RAW264.7 cells. As expected, M@SINPs treatment could potentially repolarize RAW264.7 cells to M1 phenotype with the highest M1-biomarkers expression levels and the lowest M2-biomarker expression levels among all treatment groups.

In vivo antitumor efficacy of M@SINPsFig. 4

In vivo antitumor therapeutic evaluations of M@SINPs. (A and B) Tumor volume changes of mice after the indicated treatments (n = 5). (C-E) FCM analysis of the intratumoral M1 macrophage (iNOS+), M2 macrophage (CD206+) and the ratio of M1/M2 (n = 3). (F-I) Cytokines levels in serum analyzed by ELISA kits (n = 3). (J and K) FCM analysis of the intratumoral CD8+ T cells and CD4+ T cells (n = 3). (L) The ratio of CD8/Treg in the tumor form the treated mice (n = 3). (M) Representative immunohistochemical images of tumor sections showing the infiltration of CD206+, iNOS+ and CD8 + cells (scale bar = 20 μm). DAPI was used to stain the nucleus of the cell (blue). (#< 0.05, ## < 0.01, ### < 0.001; * vs. PBS, *p < 0.05, **p < 0.01, ***p < 0.001)

Encouraged by the results from our in vitro studies, the antitumor ability of M@SINPs was further evaluated in CT26-bearing mice. The mice were i.v. injected with all formulations, and then under laser irradiation at 12 h post-injection as described in the Methods section. Compared to PBS group, SNPs, SINPs and M@SINPs group without laser irradiation groups induced a significant tumor growth inhibition (Fig. 4A and B). Previous reports have also demonstrated the effect of SHP099-mediated SHP-2 inhibition in controlling tumor growth in vitro and in vivo, which are consistent with our study [38,39,40]. More importantly, combination therapy with M@SINPs and laser irradiation (M@SINPs + Laser group) showed the greatest antitumor efficiency, attributed to the mannose-mediated targeted drug delivery and the therapeutic performance by simultaneous phagocytosis and polorization macrophages. Additionally, no apparent body weight loss and no obvious histopathological organ damage were observed in the M@SINPs-treated mice, suggesting no apparent toxicities caused by i.v. injected M@SINPs (Figure S7 and S8).

To validate the antitumor immune response induced by M@SINPs, immune cell populations at the tumor sites were firstly tested. M1 macrophages is crucial to stimulate the infiltration of intratumoral CD8+ cytotoxic T lymphocytes (CTLs) and hinder the immunosuppressive function of regulatory T cells (Tregs) [41]. Herein, the polarization effect of M@SIPNs were evaluated using FCM in the CT26-bearing mice after the various treatments. We found that intratumoral TAMs infiltration was significantly influenced by the SINPs or M@SINPs with laser treatment, with both exhibiting higher percentage of M1 macrophage (CD45+F4/80+iNOS+) and lower percentage of M2 mcarophage (CD45+F4/80+CD206+) when compared to SINPs or M@SINPs without laser irradiation (Fig. 4C and D). As a result, SINPs + Laser or M@SINPs + Laser treatment significantly increased the M1: M2 macrophage ratio when compared with all other groups (Fig. 4E). We examined the intratumoral ROS production in tumors of mice i.v. administrated with SINPs + Laser or M@SINPs + Laser. Figure S9 showed that compared with PBS group, obvious ROS production (green fluorescence) could be observed on the tumor slices collected from the mice of SINPs + Laser or M@SINPs + Laser treatment group, respectively. The above results indicated that under the laser irradiation, SINPs or M@SINPs could generate ROS at tumor sites, which could promote TAMs M2-to-M1 polarization in tumor site. Moreover, the highest M1-to-M2-macrophges ratio was found in the M@SINPs + Laser group (2.27%), which was to be 4.1- and 2.0-fold higher than those of M@SINPs without laser irradiation group (0.56%) and SINPs + Laser group (1.16%), respectively, implying that M@SINPs + Laser treatments significantly promote the re-education of M2-TAMs. Additionally, as illustrated in Fig. 4F and I, the level of M1-associated cytokines including TNF-α and IL-6 in the M@SINPs + Laser group were upregulated by 6.85- and 13.3-fold, respectively, compared to the PBS group, whereas M2-associated cytokines including IL-10 and TGF-b were downregulated. Simultaneously, the most CD8+ T cell and CD4+ T cell infiltration in tumors were observed in the M@SINPs + Laser group (Fig. 4J and K, S10). Moreover, M@SINPs + Laser treatment increased the CD8+ T cells/Tregs ratios by 7.9 times compared with the PBS group, suggesting the reversal of the immune-suppressive microenvironment (Fig. 4L). The intratumoral infiltration of CD206+, iNOS+, and CD8+ cells was further validated by immunofluorescence staining. M@SINPs + Laser group resulted in the lowest CD206 expression and the highest iNOS and CD8 expression, which was consistent with the FCM analysis results in all treatment groups (Fig. 4M). Taken together, these data provide evidence of the remodulation of tumor immune microenvirionment by M@SINPs + Laser treatment, which apparently contributed to the most effective inhibition of tumor growth.

M@SINPs improved efficacy of aPD-1-block immunotherapy

TAMs repolarization towards the antitumor M1 phenotype is crucial for improving the therapeutic efficacy of checkpoint inhibitor therapy such as anti-PD-1 antibody (aPD-1) blockade [42, 43]. In this study, we then investigated whether the M@SINPs + Laser treatment in combination with aPD-1 could induce the synergistic antitumor effect. CT26-bearing mice were injected with either PBS, M@SINPs + Laser, aPD-1, or a combination of M@SINPs + Laser and aPD-1. As expected, M@SINPs + Laser treatment showed synergy with aPD-1 in inhibiting tumor growth, and significantly extended the survival of the treated mice (Fig. 5A-C). By analysis of the infiltrating immune cells in tumor tissues after different treatments, we found that the combination-treated tumor had the most intratumoral infiltration of CD8+ T cells (17.58%) and M1 macrophages (iNOS+, 33.93%), and in turn, the lowest intratumoral infiltration of Tregs (6.16%) and M2-TAMs (CD206+, 8.58%) (Fig. 5D-G). As a result, both the CD8/Treg ratio and the M1/M2 ratio in the tumor with the combination treatment increased by more than 4-fold of that in the tumor treated with aPD-1, demonstrating the conversion of “cold tumors” to “hot tumors” (Fig. 5H and I). Additionally, the elevated pro-inflammatory cytokines levels were detected in the combination treatment group, as shown by a 2.9-fold increase in IL-6 and a 1.8-fold increase in TNF-α compared with the aPD-1 group, respectively (Fig. 5J and K). All these data suggested that M@SINPs + Laser treatment improved the antitumor efficacy of the aPD-1 blockade.

Fig. 5

In vivo therapeutic efficacy of M@SINPs in combination with PD-1 blockade. (A-C) Antitumor effects in terms of tumor growth and survival in CT26-bearing mice with treatment of either PBS, M@SINPs + Laser, aPD-1, or a combination of M@SINPs + Laser and aPD-1 (n = 5). (D-G) Representative FCM plots and quantitative analysis of intratumoral CD8+ T cells, Tregs, M1 macrophage (iNOS+) and M2 macrophage (CD206+) (n = 3). (H and I) The ratio of CD8/Treg and M1/M2 in the tumor form the treated mice (n = 3). (J and K) Cytokines levels in serum analyzed by ELISA kits (n = 3). (#< 0.05, ## < 0.01, ### < 0.001; * vs. PBS, *p < 0.05, **p < 0.01, ***p < 0.001)