Biodegradable nanoparticles-mediated targeted drug delivery achieves trans-spatial immunotherapy

Abstract

Immunotherapy has been seriously retarded due to inadequate antigen presentation and a tumor cell-dominated immunosuppressive microenvironment (TME). Herein, biodegradable multifunctional mesoporous silica nanoparticles, with dispersed carbon nanodots incorporated into the frameworks, active TKD peptide modification on the surfaces and hydrophobic drug loading in the pores, were prepared for targeted chemotherapy synergized with trans-spatial immunotherapy. The nanoparticles were biodegradable due to nanodot-induced framework swelling, which would (1) kill the in situ tumor cells and promote antigen release by targeted chemotherapy and (2) trigger biodegraded debris involving TKD and CDs to largely adsorb the tumor antigens via π-π conjugation synergized hydrophobic interactions and then massively transport these antigens from the tumor cell-dominated TME to the immune cell-dominated spleen via TKD-mediated small size effects. Thereafter, these antigens can be processed into antigen peptides via TKD-mediated lysosome endocytosis and then activate T cells in the spleen via MHC complex construction and dendritic cell cytomembrane presentation. Therefore, improved immunotherapy with trans-spatial antigen presentation avoided TME immunosuppression, which when synergized with targeted chemotherapy, markedly enhanced the therapeutic outcomes of triple-negative breast cancer.

Graphical abstract

1. Introduction

Immunotherapy has been regarded as a promising cancer treatment due to its great potential in inhibiting recurrence/metastasis, overcoming multidrug resistance and lowering side effects [1], [2], [3]. Commonly, the antitumor immune response for immunotherapy includes: (1) initial release and presentation of tumor antigens, (2) subsequent activation and migration of tumor-targeted effector T cells, (3) effective infiltration of immune cells into deep tumor tissue with (4) specific recognition and removal of tumor cells. Thus, antigen presentation and effector T-cell activation not only initiate immunotherapy but also induce immune escape [4], [5]. However, inadequate tumor antigens and their inefficient presentation in line with a high percentage of T_regs with overexpression of PD-L1 as well as other immune inhibition factors in the tumor microenvironment (TME) seriously impede the activation of T_effs and depress immunotherapy [6], [7]. In contrast, various strategies, including immunogenic cell death (ICD) [8], immune checkpoint inhibition [9] and T-cell reformation [10], have been adopted to increase tumor antigen generation and elevate T-cell stimulation to improve immunotherapy. However, ICD requires complicated treatments, where even so, the produced antigens can still not be well exposed [11]. Meanwhile, immune checkpoint inhibition is often based on a single target, which cannot overcome multisite immune inhibition in tumor cell-dominated TMEs [12]. Moreover, in vitro T-cell reformation usually suffers from complex steps with high costs (low yields) and low specificity with limited immune cognition, which causes an unacceptable curative effect on special heterogeneous tumors [13]. Comparatively, in vivo in situ antigen presentation (generation, exposure and recognition) and consequent activation of T cells in immune cell-dominated organs (such as the spleen) away from the TME, would be an efficient strategy to overcome the deficiencies of current immunotherapy since it is a simple tumor antigen-activated multisite immune response mediated by ICD. Nevertheless, this strategy requires adequate targeted transportation of tumor-associated antigens (TAAs), especially tumor-specific antigens (TSAs), from the TME into immune organs with subsequent efficient antigen processing to activate T cells based on a simple targeted therapy-induced ICD. Therefore, a powerful therapeutic system is urgently needed.

Doxorubicin (DOX)-based chemotherapy, a simple cancer treatment, has shown potential for inducing tumor antigen production, even when immunogenicity is inadequate; thus this property could be exploited to improve immunotherapy through combination treatments [14]. Meanwhile, TKD peptide (HSP70 peptide) [15], the specific ligand of HSP70 receptors, could mediate nanoparticles to target tumor cells [16]. Simultaneously, this peptide might be capable of mediating nanoparticles to enter CD91-overexpressing DCs since HSP70 is also a highly affinitive targeted ligand of CD91 [17], which might lead adsorbed antigens to be transported into DCs in the spleen and serve as guidance to escape from the immunosuppressed TME. More importantly, CD91-mediated DC uptake increases lysosome endocytosis and improves antigen processing from proteins to peptides in DC-dominated immune organs to enhance T-cell activation [18]. Additionally, owing to the hydrophobicity and aromatic structure of most tumor antigens [19], [20], small nanoparticles with large surface areas and hydrophobic π-conjugated frameworks would benefit the delivery and exposure of tumor antigens.

Therefore, in this work, biodegradable multifunctional mesoporous silica nanoparticles (DCMPT) with dispersed carbon nanodots incorporated into the frameworks, active TKD peptide linked on the surfaces and hydrophobic DOX loaded in the pores were prepared for targeted chemotherapy synergized with immunotherapy of triple-negative breast cancer (Fig. 1). The multifunctional nanoparticles precisely delivered DOX to kill in situ tumor cells and promote antigen exposure. Meanwhile, these nanoparticles could be biodegraded via nanodot-induced framework swelling to release the incorporated carbon nanodots along with TKD-containing debris, which could largely adsorb the tumor antigens via π-π conjugation synergized hydrophobic interactions and then targeted transport these antigens from the tumor cell-dominated immunosuppressed TME into the immune cell-dominated spleen for trans-spatial antigen presentation due to the relatively small sizes and special TKD mediation. Thereafter, these antigens could selectively enter the lysosome to be processed into antigen peptides via TKD mediation, which thus could be combined with MHC to construct the MHC complex and then present them on DC cytomembranes to activate T cells. As a result, improved immunotherapy with desirable therapeutic outcomes was achieved, which synergized with targeted chemotherapy to remarkably inhibit tumor growth and recurrence and significantly prolong the survival time of triple-negative breast tumor-bearing mice.

Fig. 1.

 Schematic illustration of DCMPT for enhanced immunochemotherapy via targeted drug delivery synergized trans-spatial antigen presentation.

2. Materials and methods

2.1. Materials

TKD (CTKDNNLLGRFELSG) was synthesized by Ziyu Biotech Co., Ltd. (Shanghai, China). Cyanine 5 (Cy5). Doxorubicin (DOX), cell counting kit-8 (CCK-8) and phosphate buffered saline (PBS) were obtained from Meilunbio Co., Ltd. (Dalian, China). N-hydroxylsuccinimide (NHS)-polyethylene glycol-maleimide (NHS-PEG-MAL, MW 3.5k) was bought from JenKem Technology Co., Ltd. (Beijing, China). Olorimetric TUNEL apoptosis assay kit was purchased from Beyotime Co., Ltd. (Shanghai, China). Graphene carbon nanodots (CDs) were purchased from Xianfeng Nano (China). Other reagents such as hexadecyl trimethyl ammonium bromide (CTAB), triethylamine (TEA), 3-aminopropyltriethoxysilane (APTES) and ammonium chloride (NH₄Cl) were purchased from Sinopharm Chemical Reagent Co., Ltd (China). IDO-IN-2 was bought from MedChemExpress (U.S.A.). MCF-7 breast cancer cells, MCF-10a breast normal cells and 4T1 breast cancer cells were obtained from Shanghai Institute of Cell Biology, Chinese Academy of Sciences.

2.2. Preparation of CMPT and their counterparts

Carbon dots-incorporated mesoporous silica nanospheres (CD@MSN, CM) were prepared via surfactant-mediated co-self-assembly of silica precursor and polymer-coated carbon dots. Typically, 2 mg CDs with hydrophilic functional groups were dispersed in 22 ml cetyltriethylammnonium bromide (CTAB)-contained water solution (8.3 mg/ml). The pH of this solution was adjusted to 9.5 via adding appropriate amount of TEA. Then, a mixture of 1.1 mmol TBOS (Butyl orthosilicate) and 0.1 mmol (3-aminopropyl) triethoxysilane (APTES) were added to the solution and stirred at 80 °C for 6 h. The precipitate was recovered by centrifugation (12,000 rpm, 5 min) and repeated water/ethanol washing. To remove the residual CTAB, the product was stirred in NH₄NO₃ ethanol solution (0.6 wt%) at 70 °C for 3 h. Finally, CMs was obtained by freeze-drying.

Accordingly, undegradable MSNs were prepared without using carbon dots. Cy5-labeled nanoparticles were synthesized using the preprepared Cy5-APTES solution (0.01 mmol of Cy5-NHS ester and 0.011 mmol APTES were stirred in anhydrous DMSO under a dark and oxygen-free atmosphere for 2 h) instead of APTES during the same synthesis. This preparation basically had the same fluorescence labeling efficiency for different nanoparticles, where accordingly Cy5-labeled CDs with similar fluorescence intensity were also prepared via stirring the appropriate amount of Cy5-NHS ester and CDs in DMSO.

Furtherly, PEGylation were achieved by stirring 0.2 mg CM or its counterparts (2 mg/ml) and 0.4 ml functionalized MAL-PEG-NHS (1 mg/ml) in PBS 8.0 for 2 h. Then, TKD was linked on the nanoparticles via click reaction between 0.2 ml TKD (HSP70) (0.5 mg/ml) and the PEGylated nanoparticles in PBS 7.4. All these nanoparticles were purified via centrifugation and water-washing.

2.3. Characterizations

Thermogravimetric (TG) spectrum was acquired by Germany SDT Q600-GSD 301T2. TEM, STEM, HRTEM and EDS-mapping images were performed by using Talos F200X TEM (ThermoFisher, U.S.A) combined with Gatan software. Infrared (IR) spectra were carried on Thermo Nicolet 360FT-IR using KBr pellet method. X-ray photoelectron spectroscopy (XPS) was measured by an RBD-upgraded PHI-5000C ESCA system. Dynamic light scattering (DLS) and Zeta Potentials were analyzed with zeta potential/particle sizer Malvern 3600 (U.K.). UV–visible spectra and FL spectra were obtained on UV2401PC (Shimadzu, Japan) and LS-55 (PerkinElmer, U.S.A.), respectively.

2.4. Drug loading and release

To loading DOX into the nanoparticles, 0.5 ml DOX aqueous solution (1 mg/ml) was added to the above CMPT or its counterpart solution and stirred at room temperature for 8 h. After removing the loosely absorbed DOX on the surface of nanoparticles via centrifugation and water-washing, the DOX-loaded nanoparticles were obtained. The DOX loading efficiency (LE) and entrapment efficiency (EE) were calculated by the followed functions. The drug content was quantified by fluorescence intensity of DOX (Perkin Elmer LS-55). To evaluate drug release, DOX-loaded nanoparticles were shaking in PBS solutions with different pH value. At certain time intervals, the released drug was determined by fluorescence intensity with pre-established standard curves.

$$LE{(\%)}_{loadingefficiency}=M_{loadeddrug}/M_{nanoparticles}$$

$$EE{(\%)}_{entrapmentefficiency}=M_{loadeddrug}/M_{totaladdeddrug}$$

2.5. Evaluation of biodegradation

The nanoparticles were shake in simulate TME solutions (PBS 6.0) at 37 °C for different time periods. At a certain time-point, the debris were recovered via centrifugation and analyzed via particle sizer/zeta potentials, TEM images, UV–vis absorption spectra and so on. The instrument used was mentioned above.

2.6. Cell culture

Bone marrow-derived dendritic cells (BMDC) were extracted and induced to differentiation according to the existing maturation scheme. Typically, bone marrow cells of Balb/c mice were washed out from the marrow using 0.5 ml DCs culture medium (RPMI-1640 media, 10% FBS, 1% penicillin, 1% l-glutamine,1% streptomycin, 10 ng/ml IL-4 and 20 ng/ml GM-CSF). Then, the obtained marrow cells were cultured in 10 ml DCs culture medium, and every two or three days, half of the medium (5 ml) was renewed. After 7-day culture, the semi-adherent cells were collected for staining and flow cytometry analysis. MCF-7 and 4T1 cells were maintained in tumor cells culture medium (RPMI-1640 media, 10% FBS, 1% l-glutamine, 1% streptomycin, and 1% penicillin) and cultured in a humidified environment of 5% CO₂ at 37 °C. The medium was renewed every two or three days with cells being trypsinized before plating.

2.7. Animal model

All the mice were raised in Experimental Animals Department of Fudan University. The procedures for care and use of animals were approved by Ethics Committee of Fudan University with all applicable institutional and governmental regulations concerning the ethical use of animals being followed. When the female Balb/c mice were weighted about 20 g in the absence of specific pathogens, 1 × 10⁶ 4T1 cells were subcutaneously injected into the right shoulder to establish breast cancer-bearing mice models.

2.8. Evaluation of targeting uptake in cells

MCF-7 cells or MCF-10 cells (8 × 10³ per well) were first cultured in glass bottom cell culture dished (Corning, 96 wells) with 100 µl of media for 24 h. Then, 100 µl DCMPT and DCMP (200 µg/ml in CMP) were added into the dishes and incubated for 5 min, 15 min, 30 min, 1 h and 2 h, respectively. Correspondingly, the fluorescence was detected by fluorescence inverted microscope (OLYMPUS, CKX53) to indicate cell uptake.

2.9. Evaluation of cytotoxicity

MCF-7 cells or MCF-10 cells (8 × 10³ per well) were incubated with DCMPT with different concentrations from 0 ug/ml to 500 ug/ml for 2 h, and then the cell viability was detected by CCK-8 (Meilunbio MA0218).

2.10. Investigation of the uptake pathway

To investigate the uptake pathway of nanoparticles by tumor cells and DCs, MCF-7 cells, 4T1 cells or DCs (acquirement mentioned in Cell Culture) in the density of 1.5 × 10⁵ per well were cultured in glass bottom cell culture dishes (NEST, 15 mm) with 1.5 ml media for 24 h. After rinsing with PBS, 1.5 ml DCMPT or DCMP (containing 200 µg/ml CMP) were added into the dishes and incubated for another 5 min, 15 min, 30 min, 1 h and 2 h, respectively. Lysosomes and nucleus were labelled via lysotracker green and DAPI. Then cells were washed three times with PBS. After that, the cells were observed in fresh medium by confocal microscopy (Carl Zeiss LSM710).

2.11. Proliferation and differentiation of DC

The experimental process is shown in Fig. S12. In short, 2 × 10⁶ 4T1 was incubated with different treatments for 48 h (Saline, CMPT (the same concentration of CMPT in DCMPT), DMPT (the same concentration of silica nanoparticles in DCMPT) and D&I (the same concentration of DOX in DCMPT, manufacturer instructed concentration of IDO-IN-2), DCMPT (2 ml, 200 µg/ml)). Then, the cells supernatants were centrifuged (1200 rpm, 5 min) to remove the cells, which was followed by shaking for 96 h and centrifugation (12,000 rpm, 5 min). After that, the supernatants were skimmed to remove weakly adsorbed protein. The precipitate was dissolved in 1 ml PBS. Finally, the as-obtained solution was incubated with 7 × 10⁵ primary bone marrow-derived dendritic cells for 24 h (BMDC were acquired according to the well-established DC maturation protocol mentioned above in Cell Culture and Animal Model). Then, the expression of CD86 (stained by ThermoFisher PA5–79,007) / CD11c (stained by ThermoFisher PA5–110,685)/ MHCI (stained by ThermoFisher 11–5998–81) or MHCII (stained by ThermoFisher 36–5321–85) on dendritic cells were measured by flow cytometer (BD FACAria U.S.A.).

2.12. Adsorption of tumor associated antigen and tumor specific antigen

This experimental process is shown in Fig. S12. In short, 2 × 10⁶ 4T1 cells were incubated with different administration groups for 48 h (saline, CMPT (the same concentration of carbon dots in DCMPT), DMPT (the same concentration of silica nanoparticles in DCMPT) and D&I (the same concentration of DOX in DCMPT, manufacturer instructed concentration of IDO-IN-2), DCMPT (10 ml, 200 µg/ml)). Then, the supernatants were centrifuged (1200 rpm, 5 min) to remove the cells, which was followed by shaking for 96 h and centrifugation (12,000 rpm, 5 min). After that, collected supernatant was skimmed to remove the weakly adsorbed proteins. Finally, the precipitate was dissolved in 1 ml PBS, and the total amount of protein was determined by BCA Protein Assay Kit (Meilunbio MA0082).

2.13. Pharmacokinetic of DOX and distribution of nanoparticles in vivo

To explore the pharmacokinetic properties of DOX, female Balb/c mice bearing tumor around 100 mm³ were divided into three groups, DOX (pure DOX, 100 µg/ml), DCMP (the same concentration of DOX) and DCMPT (the same concentration of DOX), with three mice in each group. After intravenous injection, the whole blood (500 µl) from eye sockets of mice were collected at different post-injection time points. After standing for 30 min, the obtained whole blood was centrifuged at 4 °C before 300 µl supernatant being collected as plasma. The DOX concentration was calculated by measurement of its fluorescence intensity (FL) according to the standard curves. The pharmacokinetics of DOX were determined using two-compartment models due to the abnormal blood flow in tumor. Typically, the curves were recorded via the logarithm of blood drug concentration vs time points post-injection. Correspondingly, distribution and elimination phase in this curve were linearly fitted, respectively, to obtain the distribution and elimination rate.

2.14. In vivo distribution of nanoparticles and their debris

To demonstrate the in vivo distribution of different nanoparticles and their debris. Female Balb/c mice bearing tumor around 100 mm³ were divided into three groups, C-cy5 (Carbon Dots, the same concentration of carbon dots in CMPT-cy5), MPT-cy5 (the same concentration of silica nanoparticles in CMPT-cy5) and CMPT-cy5 (10 ml/kg, 200 µg/ml), with three mice in each group. At different time points after intravenous injection, the FL intensity of mice was detected by in vivo fluorescence imaging system (Visque Invivo Smart, Koeran). At different periods (8 h, 1 d, 3d, 6 d, 9 d and 12 d) post-injection (maximal FL intensity at tumor sites), the mice were sacrificed and major organs together with spleens and tumor draining lymph nodes were taken for FL imaging.

2.15. In vivo accumulation in spleen and tumor

Female Balb/c mice bearing tumor around 100 mm³ was divided into 4 groups: MPT I.V. (the same concentration of silica nanoparticles in CMPT I.V.), MPT I.T. (the same concentration of silica nanoparticles in CMPT I.T.), CMPT I.V. (10 ml/kg, 200 µg/ml), CMPT I.T. (10 ml/kg, 200 µg/ml). Here, I.V. is the abbreviation of intravenously injection and I.T. is the abbreviation of intratumoral injection. At different time points from 8 h (0.33d) to 12 days after different treatments, the spleens and tumors of mice were taken out, crushed and then treated with aqua regia (65% HNO₃ : 36% HCl = 1 : 3) for 12 h to be completely dissolved. Consequently, the silicon element is measured by ICP-OES (Perkinelmer Optima 8000).

2.16. Hemocompatibility tests

The blood compatibility of prepared nanoparticles was evaluated by Hemocompatibility Tests. The experiment used heparin-stabilized 4% goat red blood cell suspension was bought from Baisiyi Co., Ltd (China). The test groups were acquired by adding 100 µl suspension to 900 µl nanoparticles solution with different concentration (100 µg/ml, 200 µg/ml, 400 µg/ml)_. The negative control group was acquired by adding 100 µl suspension to 900 µl PBS while the positive control group was acquired by adding 100 µl suspension to 900 µl deionized water. All groups were standing for 2 h and then were analyzed by UV2401PC after centrifugation (10,000 rpm, 1 min). The hemolytic rates (HR) of different nanoparticles and drugs were calculated by the followed functions.

$$HR{(\%)}_{HemolyticRate}=\frac{D_t-D_{nc}}{D_{pc}-D_{nc}}\times 100\%$$

D_t, D_nc and D_pc were OD values at 540 nm of test group, positive control group and negative control group, respectively.

2.17. Bio-degradation mediated antigen adsorption

In this experiment, 2 × 10⁶ 4T1 cells were lysed to obtain tumor cell lysate with protein concentration of 10 mg/ml (tested by BCA Protein Assay Kit (Meilunbio MA0082)). Different groups of CD (the same weights of carbon dots in CMPT), MPT (the same weights of silica nanoparticles in CMPT), and CMPT (200 ug/ml) were incubated with tumor cell lysate for different time periods. Then, the as-obtained cell lysates were centrifuged (12,000 rpm, 5 min) to remove weakly adsorbed proteins. After this, the precipitate was dissolved in 1 ml PBS, where the total adsorbed proteins were determined by BCA Protein Assay Kit (Meilunbio MA0082) with specific TAAs analyzed by Western blot.

2.18. In vivo therapeutic effect

To evaluate the therapeutic outcomes, mice on the 7th day after inoculation with 4T1 tumor cells (tumor volume: around 100 mm³) were injected intravenously with saline, CMPT (the same concentration of CMPT in DCMPT), DMPT (the same concentration of silica nanoparticles in DCMPT) and D&I (the same concentration of DOX in DCMPT, manufacturer instructed concentration of IDO-IN-2), DCMPT(10 ml/kg, 200 µg/ml), respectively. Then, their weights and survival rate were recorded. On the 21st day after injection, 5 mice of each group were sacrificed with main organs and whole tumors being taken out for in vivo toxicity evaluation via TUNEL detection and hematoxylin & eosin (H&E) analysis, where the damage of each tissue was observed under the Leica DMI4000 D fluorescence microscope (Germany). Meanwhile, the lungs, tumors and spleens from mice with different treatments were imaged via a photo camera. Besides, spleens and lungs of another three mice in each group were collected for Evaluation of In Vivo Immune Effects (Materials and Methods 2.19). Blood of these three mice was reserved for Cytokine Detection (Materials and Methods 2.20).

2.19. Evaluation of in vivo immune effects

To investigate the immune effects, numbers of lung nodes and weights of spleens in each group were recorded. To analyze the proliferation and differentiation of immune cell, lymphocytes cells were obtained from the spleens of mice mentioned in In Vivo Therapeutic Effect (Materials and Methods 2.18). According to the instructions of manufacture, the lymphocytes cells were blocked with 10% bovine albumin (BSA) solution and was stained with CD348 (ThermoFisher CD3–4–8-A). The CD3, CD4, CD8-expressed cells were detected by flow cytometry (BD FACAria U.S.A.).

2.20. Cytokine detection

Whole blood was obtained from the mice in In Vivo Therapeutic Effect (Materials and Methods 2.18). First, 1 ml blood was taken out from the eye sockets of the mice. After standing for 30 min, the blood was centrifuged at 4 °C (3000 rpm, 15 min) to remove the precipitate. Then, 300 µl of the supernatant was collected as plasma for TNF-α (stained by ThermoFisher CHC1753), TNF-β (stained by ThermoFisher BMS202TEN) and IFN-γ (stained by ThermoFisher MHCIFG05) quantification via ELISA analysis according to the manufacturer's instructions.

2.21. Tumor recurrence inhibition

Tumor recurrence inhibition was investigated after surgical resection, where the treatment groups, mice models and administration dose were same as those in In Vivo Therapeutic Effect (Materials and Methods 2.18). Typically, 1 × 10⁶ 4T1 cells was injected subcutaneously into mice on the 0th day for tumor growth. On the 7th day (tumor volume: around 100 mm³), the mice were intravenously injected with different nanoparticles. On the 14th day, the tumors were surgically removed with repeated injection. Meanwhile, 1 × 10⁵ 4T1 cells was injected intravenously into mice as CTCs (circulating tumor cells). On the 21st day, the mice were intravenously injected with different nanoparticles again. During this process, the weights and survival rates of the mice was recorded. On the 23rd day, five mice of each group were sacrificed for lung nodes counting and spleen weights recording. Correspondingly, H&E stained for lungs were performed for tumor recurrence evaluation. Another five mice of each group were continued for survival rate determination. Additionally, on the 23rd day, 1 ml blood was taken out from the eye sockets of the mice. After standing for 30 min, the blood was centrifuged at 4 °C (3000 rpm, 15 min) to remove the precipitate. Then, 300 µl of the supernatant was recovered and incubated with 2 × 10⁶ 4T1 cells or MCF-7 cells for 2 h. Accordingly, the cell-mediated cell toxicity was detected by Live-Dead imaging using ThermoFisher L7010 according to the manufacturer's instructions. The scheme was shown in Fig. 6a.

Fig. 6.

Tumor recurrence inhibition. (a) Scheme of the different treatments. (b) Photos, (c) representative H&E-stained images and (e) node numbers of lungs harvested from the mice on 21st day after injections of different nanoparticles. The red dotted circles in (b) and the red lines in (c) showed the palindromic tumors and the tumor borders in lungs, respectively (Bar = 200 µm). (d) CLSM images of 4T1 tumor cells incubated with the blood from the mice on 21st day after injections of different nanoparticles (Bar = 50 µm). (f) Body weight and (g) Kaplan-Meier survival curves of the recurrence-modeled mice treated by different nanoparticles. Data were represented as mean ± SD (n = 5). Statistical significance was calculated by one-way ANOVA using the Tukey post-test (**p < 0.01).

3. Results and discussion

A general surfactant-mediated co-self-assembly of silica precursors and polymer-coated carbon nanodots was developed to prepare biodegradable mesoporous silica nanoparticles (Fig. S1) [21]. Here, red fluorescence-emitting CDs with larger π-conjugated planes could be incorporated (Fig. S2 and Fig. S3), which, cooperatively with aminated silica precursors (APTES & TBOS), derived biodegradable amido-terminated mesoporous silica (CD@MSN—NH₂, CM) with potentially enhanced π-adsorptions. Typically, CMs were uniform nanoparticles with dispersed and crystalline carbon nanodot incorporation (approximately 10 wt% from TG analysis, Fig. S4), as observed from TEM, HRTEM and the corresponding Fourier transformed images (Fig. 2a-d). Then, bifunctional PEG (NHS-PEG₃₄₀₀-MAL) was linked to CD@MSN—NH₂ (CMP) via the specific reaction between -NH₂ and -NHS, which was followed by covalent TKD conjugation to yield TKD-modified nanoparticles (CMPT) via the reaction between -SH of TKD peptide and -MAL of PEG (Fig. S3). The successful synthesis of CMPT was validated by TEM, STEM and EDS-mapping images, where CMPT possessed an extra N element originating from TKD and relatively more asperous surfaces than CM (Fig. S5 and Fig. 2e,f). Meanwhile, the decreased zeta potentials, the gradually increased hydrated particle sizes (Fig. 2g) and especially the emerged TKD absorbance as well as the obvious C, O, N and Si elements in CMPT also suggested PEG and TKD linkages to CM via the abovementioned reactions (Fig. S6). Moreover, the reserved C—O stretching vibration (1400 cm⁻¹) from CDs in CM, the emerged C—H stretching vibration (-CH₂, 2800 cm⁻¹) and disappeared N—H stretching vibration (-NH₂, 3400 cm⁻¹) in CMP, and then the enhanced O—H (3650–3200 cm⁻¹)/N—H (3500–3100 cm⁻¹)/C—N (1200 cm⁻¹) stretching vibrations in CMPT further verified the adequate CDs incorporation, the covalent PEG linkage and the targeting TKD modification, respectively (Fig. 2h). This modification endowed CMPT with improved water dispersity (Fig. S7). More importantly, being different from the pure MSN (Fig. S8), the as-synthesized CMPT was degradable in the simulated TME (PBS 6.0) (Fig. 2k-o) owing to the relatively low energy Si-C bonds (Fig. 2i and Fig. S6) produced by CD incorporation into the framework of MSN, where the CDs-involved nanodebris associated with the reduced particle sizes and decreased zeta potentials (Fig. 2j) was gradually produced. In particular, the reserved TKD absorbance in the degraded products also suggested the involvement of the TKD peptide in the debris in addition to the CDs during biodegradation (Fig. S9). Since CDs possessed hydrophobic-π-conjugated planes [22] and TKD could target HSP70 and CD91 overexpressed on tumor cells and DCs, respectively, these CDs and TKD-involved debris could augment the bioapplications of CMPT.

Fig. 2.

Characterizations of multifunctional biodegradable DCMPT. (a) TEM image, (b, c) HRTEM images and (d) inverse Fourier transformed (FT) HRTEM image of CM. The inset in (c) and (d) showed FT patterns and lattice fringes of the incorporated CDs into CM, respectively. (e, f) STEM images and EDX mappings of silica, oxygen, nitrogen and carbon for CM and CMPT, respectively. (g) Particle sizes, zeta potentials and (h) FT-IR spectra of CD, CM, CMP and CMPT, respectively. (i) High-resolution C 1 s XPS spectra of CMPT. (j) Particle sizes and zeta potentials of CMPT with different degradation time periods. (k-o) TEM images of CMPT degraded for 1, 3, 6, 9 and 12 days, respectively. Data were represented as mean ± SD (n = 3).

Initially, due to its large surface area and appropriate pore sizes, CMPT could load the anticancer drug DOX via π-π and hydrophobic interactions with a maximal entrapment efficiency of 90.2% and a saturated loading efficiency of 47.4% with stable particle size (Fig. 3a and S10). Furthermore, owing to TKD mediation, CMPT could deliver much more drug into HSP70-overexpressing MCF-7 tumor cells (Fig. 3b) than CMP, but neither CMPT nor CMP could deliver the drug into MCF-10A normal cells effectively. Meanwhile, DOX-loaded CMPT (DCMPT) exhibited a pH-responsive drug release (Fig. 3c), the rate of which was significantly elevated in acid solution due to the reduced hydrophobic π-π interaction between DOX and graphitic CDs and the enhanced electrostatic repulsion between protonated DOX and electropositive MSN [23]. This approach was beneficial for tumor-targeted chemotherapy because of the acidic tumor microenvironment. In particular, TKD mediation also increased the lysosome endocytosis of the nanoparticles; therefore, DCMPT showed a more obvious drug accumulation in the acidic lysosomes of tumor cells than DCMP (Fig. 3e). In this situation, DCMPT induced much more significant cytotoxicity for tumor cells than for normal cells because of the TKD-mediated targeting uptake and consequently the acid-triggered drug release in lysosomes (Fig. 3d). In addition, owing to the specific high affinity of HSP70 with CD91 overexpressed on DCs [17], HSP70 peptide (TKD) mediation could also increase DC uptake and lysosomal endocytosis of nanoparticles, as verified by the much more intensive fluorescence signals in DCMPT-incubated tumor cell lysosomes than in DCMP-incubated ones, especially under prolonged incubation time periods (Fig. 3f). These results suggested targeted drug delivery into tumor cells through DCMPT and consequently its biodegradation into CDs and TKD-involved debris with DC targeting ability. Additionally, to facilitate the following immunity investigations, the tumor-targeted drug delivery was also examined on rat-derived 4T1 cells (triple-negative breast tumor cells) and similar performances were observed, confirming the TKD-mediated targeting ability (Fig. S11).

Fig. 3.

Tumor cells targeting and DC activation. (a) DOX loading and entrapment efficiency curves for CMPT. Data were represented as mean ± SD (n = 3). (b) CLSM images of MCF-7 tumor cells and MCF-10A normal cells incubated with DCMP or DCMPT for different time periods. Red: DOX. Scale bar = 100 µm. (c) Cumulative DOX release curves from DCMPT in PBS solutions with different pH values. Data were represented as mean ± SD (n = 3). (d) Cell viabilities of MCF-7 and MCF-10A cells incubated with different concentrations of DCMPT for 2 h. Data were represented as mean ± SD (n = 3). (e) CLSM images of MCF-7 cells incubated with DCMP or DCMPT for different time periods. Blue: DAPI; Green: Lysosome; Red: DOX. Scale bar = 10 µm. (f) CLSM images of DCs incubated with DCMP or DCMPT for different time periods. Blue: DAPI; Green: Lysosome; Red: DOX. Scale bar = 10 µm. (g-i) Flow cytometry analysis and the corresponding quantifications of MHCI⁺CD86⁺CD11c⁺/MHCII⁺CD86⁺CD11c⁺ for DCs incubated with different TAAs-adsorbed nanoparticles. Data were represented as mean ± SD (n = 3). Statistical significance was calculated by one-way ANOVA using the Tukey post-test (*p < 0.05, **p < 0.01). (j) Concentrations of proteins absorbed on different nanoparticles with 96 h degradation after incubation with 4T1 for 48 h. Data were represented as mean ± SD (n = 3). Statistical significance was calculated by one-way ANOVA using the Tukey post-test (*p < 0.05, **p < 0.01).

Next, the recovered nanoparticles from the 4T1 tumor cells after treatment with DCMPT and its counterparts were found to have the capability to promote the maturity and differentiation of DCs (the experimental scheme is shown in Fig. S12), where both MHCI⁺CD86⁺CD11c⁺ and MHCII⁺CD86⁺CD11c⁺ cells were obviously increased (Fig. 3g-j and Fig. S13). Notably, these behaviors of DOX-loaded, TKD-modified or CD-incorporated nanoparticles were superior to their counterparts. In particular, the elevation of MHCI⁺CD86⁺CD11c⁺ and MHCII⁺CD86⁺CD11c⁺ via DCMPT treatment was more significant than the DOX-combined commercial IDO inhibitor IDO-IN-2 treatment. Correspondingly, the total protein concentrations in the nanoparticle-treated 4T1 cells were detected and were also higher for DOX-loaded or CD-incorporated nanoparticles than for their counterparts. These results suggested the efficient maturation and differentiation of DCs by nanoparticle-mediated antigen presentation, where DOX chemotherapy might induce antigen exposure. Then, biodegraded debris with a large surface area and the π-conjugated system might enhance antigen adsorption, while TKD mediation could increase the targeted uptake of the adsorbed antigens.

Then, these performances were examined in vivo via intravenous injection of the nanodrugs into 4T1 tumor-bearing mice, where DCMPT, even at a concentration of 400 µg/ml, did not cause any hemolysis (Fig. S14). First, the gradual and much more significant drug accumulation into tumor sites for DCMPT than for DCMP and especially for pure DOX confirmed tumor-targeted drug delivery via TKD mediation and its EPR effects (Fig. 4a and b). Correspondingly, DCMPT also had much longer retention within tumors than in other groups, suggesting its relatively slow metabolism via TKD and nanoparticle mediation (Fig. 4a,b and Table S1). Second, compared to free DOX, DCMP and DCMPT revealed reduced distribution coefficients and long circularity of PEGylated nanoparticle-mediated drug delivery (Fig. 4c-e and Table S1) via pharmacokinetics analysis. Accordingly, the relatively smaller elimination constant of DCMPT than that of DCMP further confirmed the TKD-mediated targeting effect, which enhanced DCMPT accumulation in tumors and thus reduced its clearance from the body (Fig. 4c-e). Third, Cy-5 was covalently conjugated on the nanoparticles instead of the non-covalently adsorbed DOX to facilitate the in vivo FL imaging of the disposition for nanoparticles (but not the DOX drug). The results showed that CMPT and its debris after biodegradation for 12 days still possessed equal FL intensity due to the stable linkage of Cy5 to the debris without release (Fig. S15). As shown in Figs. 4f and S16, the fluorescence signals in all these organs and tumors almost disappeared on the 6th day after injection of CDs, indicating their fast excretion, possibly due to their relatively small sizes. Comparatively, MPT had a much longer accumulation in the tumor. This could be attributed to its larger size and TKD-mediated tumor targeting. Nevertheless, in addition to that in the tumors, this accumulation mostly occurred in the lung as it trapped large nanoparticles (Fig. 4f). Fortunately, this lung trapping was avoided for CMPT. Instead, their retention in the spleen and tumor-draining lymph nodes was elevated and prolonged along with tumor targeting accumulation (Figs. 4f and S16). This behavior could be attributed to the biodegradation of CMPT, where TKD and CD-involved debris with small sizes could not be retained by the lung but transported to the spleen (such as DCs) via TKD mediation. In particular, in contrast to the gradual excretion of CMPT from the tumor and its distribution away from the other organs, no obvious decrease in FL in the spleen was observed on the initial injection days, which was consistent with biodegradation, implying the targeted transportation of debris from the tumor to the spleen. Meanwhile, this special transportation of debris from the tumor to the spleen was further revealed via ICP analysis for Si (Fig. 4h and i). Herein, CMPT but not MPT initially increased and then gradually decreased spleen accumulation efficiency (SAE) via both the I.V. injection and especially the I.T. injection, where the SAE at 3 days post I.T. injection of CMPT even reached a maximum of 0.57% (Fig. 4h). Correspondingly, no matter which injection was used, the tumor accumulation efficiencies (TAEs) for both CMPT and MPT gradually decreased with prolonged postinjection time periods from 0.33 (8 h) to 12 days (Fig. 4i).

Fig. 4.

In vivo distribution and tumor antigen adsorption in 4T1 tumor bearing mice. (a) Representative in vivo fluorescence images of the mice after intravenous injection of pure DOX (D), DCMP or DCMPT for different time periods (n = 3). (b) Ex-vivo fluorescence images of major organs/tumors after intravenous injection of DOX, DCMP, DCMPT for 8 h (left) and 12 h (right) post injection of nanoparticles with DOX. (c-e) Pharmacokinetics curves (Black) and the corresponding fitting curves (Red: distribution; Blue: Elimination) using two-compartment model of pure DOX (D), DCMP and DCMPT, respectively. (f) Ex-vivo fluorescence images of major organs/tumors after intravenous injection of CD-Cy5, MPT (MSN-PEG-TKD)-Cy5 or CMPT-Cy5 for different time periods. (g) Western blot and corresponding quantitation of TSAs absorbed on the different nanoparticles with degradation for 0 or 12 days. (h) Spleen (SAE) and (i) tumor (TAE) accumulation efficiency of MPT or CMPT via different injections where I.V. means intravenous injection and I.T. means intratumoral injection. Data were represented as mean ± SD (n = 3). (j) Concentrations of proteins absorbed on the nanoparticles with different degradation time periods after incubation with 4T1 cells for 48 h. Data were represented as mean ± SD (n = 3).

Hereafter, the relatively slower excretion from tumors for CMPT than that for MPT further validated the TKD-mediated targeting effects. All these results suggested the unique advantages of biodegradable CMPT, which achieved efficient tumor targeting and then enhanced transportation from the tumor to the spleen. More importantly, different from the stable high adsorption of CDs and the fixed low adsorption of MPT to tumor cell-associated proteins during the degradation treatments for different periods, the gradually increased adsorption for CMPT suggested their degradation into CD-involved debris, which could enhance protein adsorption via hydrophobic π-conjugation of the exposed CDs with aromatic amino acids from proteins (Fig. 4j). In particular, this special adsorption was also applicable to both the TAAs and TSAs (Figs. 4g, S17 and S18), where the unchanged high adsorption of CDs and the maintained low adsorption for MPT as well as the obviously elevated adsorption before and after 12 days of degradation treatment were clearly exhibited via western blot analysis. This property of DCMPT, with its advanced tumor targeting ability, enhanced transportation from tumor to spleen, and targeted chemotherapy-induced antigen production, suggested the potential tumor antigen presentation to the DC-dominated spleen for improved immunotherapy.

To confirm this, in vivo tumor inhibition and immunological effects were explored via intravenous injection of different nanoparticles into triple-negative breast tumor-bearing mice. On the 21st day after treatment, CMPT exhibited negligible tumor apoptosis, suggesting its low toxicity (Fig. 5a). After DOX loading, DMPT (no CDs incorporated and undegradable nanoparticles) induced obvious tumor apoptosis via targeted chemotherapy. Although the absence of TKD targeting would depress the chemotherapeutic effects of nanodrugs, DCMP treatment still produced apparent tumor apoptosis, revealing its biodegradation-promoted immune activation. Thereafter, DCMPT treatment significantly enhanced tumor apoptosis, which was even better than the clinical DOX chemotherapy-combined IDO inhibitor IDO-IN-2 (D&I). Correspondingly, these distinct tumor inhibitions could also be reflected by tumor volumes (Fig. 5b), tumor growth curves (Fig. 5d), tumor lesion images (Fig. 5f) and the median survival time periods of tumor-bearing mice (Fig. 5e), where DCMPT had the optimal therapeutic outcomes with minimal tumor volumes, completely inhibited tumor growth, maximal tumor lesion and over 60 days median survival. These results validated the biodegradation-promoted and TKD-mediated simultaneous targeting chemotherapy and immunotherapy for DCMPT. Meanwhile, all these treatments caused neither apparent histopathological lesions in normal organs nor obvious weight loss in tumor-bearing mice (Figs. S19 and 5c), suggesting their good biosafety. Moreover, behaving as a tumor inhibitor, these treatments also inhibited tumor lung metastasis, where the numbers of lung nodes on the 21st day after treatment were recorded as follows: control > CMPT > DMPT > DCMP > D&I > DCPMT (Fig. 5g, h, l). This suggested that immune activation synergized with DOX chemotherapy, nanoparticle degradation and TKD mediation, and thus DCPMT almost entirely inhibited tumor lung metastasis and was even superior to D&I. Notably, inhibition of splenomegaly in terms of decreased spleen weight (Fig. 5g and i) was also observed via these treatments, which was in good agreement with their effects on tumor lung metastasis inhibition. Herein, DCMPT treatment most remarkably inhibited splenomegaly and was especially more obvious than I&D treatment, suggesting that DCMPT treatment mostly enabled immunological effects to reduce splenic immune pressure [24]. Furthermore, as shown in the flow cytometric analysis of spleen lymphocytes (Fig. 5m and j), the increased CD8⁺ via treatments with IDO inhibited the chemotherapy system (D&I), degradable chemotherapy system (DCMP), and especially the TKD-mediated degradable chemotherapy system (DCMPT), confirming their inductions for immune activation with increased effector T cells (T_eff).

Fig. 5.

In vivo tumor inhibition and immune enhancement for 4T1 tumor-bearing Balb/c mice. (a) Tumor apoptosis and (b) tumor photos of mice on 21st day post-injection of different nanoparticles. Blue: DAPI-stained nucleus; green, FITC-labeled apoptosis cells. Bar = 500 µm. (c) Body weights, (d) relative tumor volume and (e) Kaplan-Meier survival curves of mice after different treatments. Data were represented as mean ± SD (n = 5). (f) Representative H&E-stained images of tumors (Bar = 200 µm), (g) photos of lungs and spleens as well as (l) representative H&E-stained images of lungs (Bar = 200 µm), (h) numbers of lung nodes and (i) weight of spleens, (m) flow cytometry analysis and (j-k) the corresponding quantified percentages of CD3⁺CD8⁺ cells and CD3⁺CD4⁺ cells in spleens, and (n) concentrations of TNF-α in plasma, from the 4T1 tumor-bearing mice on 21st day post-injection of different nanoparticles. Data were represented as mean ± SD (n = 3). Statistical significance was calculated by one-way ANOVA using the Tukey post-test (*p < 0.05, **p < 0.01, ***p < 0.001). The yellow circles in (g) and the red circles in (h) showed the pulmonary metastatic tumors.

Usually, immune activation in the TME via autoantigen presentation is restricted by simultaneously increased T_reg cells [25]. Fortunately, this DCMPT treatment did not cause an increase in CD4⁺ T_reg cells (even a slight decrease), which was obviously contrary to the results of the other treatments (Fig. 5m and k). This special immune activation via DCMPT treatment revealed its occurrence from the immunosuppressive TME to the spleen, where the antigens were presented via the first antigen-processed peptides and then its complex with MHCI and MHCII in DCs. Consequently, maturation of DCs and T_eff cells was clearly observed for the DCMPT treatment in terms of the most markedly increased secretion of antitumor chemokines, such as TNF-α, TNF-β and IFN-γ, both in plasma and tumors (Figs. 5n, S20 and S21), which would contribute to the enhanced immune killing of tumors. All these findings illuminated the TKD- and CD-involved nanodebris from DCMPT chemotherapy for efficient immune activation via targeted tumor antigen presentation from the tumor cell-dominated TME into the DC-dominated spleen.

Finally, tumor recurrence via these different treatments was examined in 4T1 tumor-bearing mice with preadministered circulating tumor cells (CTCs), where intravenous injection of the nanodrugs into the mice activated immunity, and then surgical resection of the original tumor accompanied by intravenous injection of 4T1 tumor cells and nanodrugs for recurrence observations was performed (Fig. 6a). On the 21st day after treatment, DCMPT inhibited neoplastic lung growth with almost no lung nodes, which was much superior to the other groups (Fig. 6b, c and e). Meanwhile, incubation with the blood from the mice on the 21st day after DCMPT treatments could even induce the immunogenic death of 4T1, which was much more remarkable than those of other counterparts and even the D&I (Fig. 6d). These results confirmed the advanced DCMPT-activated tumor immune killing. Moreover, owing to the inadequate immune effects acquired by CMPT, DMPT and DCMP treatments, tumor recurrence associated with CTC injection caused obvious body weight loss in triple-negative breast tumor-bearing mice (Fig. 6f). However, this drawback could be avoided by D&I and especially DCMPT treatments due to enhanced immune protection. As a result, the CTC-injected and tumor-bearing mice treated with DCMPT had the longest survival time of over 60 days among all the compared groups (Fig. 6g). All these results suggested the improved chemotherapy-synergized immunotherapy of DCMPT via their TKD-mediated targeting drug delivery and TKD/CD-containing debris-mediated efficient antigen presentation. This is a special immune activation via targeted antigen transportation from the tumor cell-dominated TME into the DC-dominated spleen followed by lysosome endocytosis and MHC complex construction, which will provide new insight into biodegradable nanoparticles for chemotherapy-synergized immunotherapy.

4. Conclusion

In summary, biodegradable multifunctional mesoporous nanoparticles, where the carbon nanodots were dispersedly incorporated into the frameworks, anticancer drug DOX was stably loaded into the mesopores and TKD peptide was covalently linked on the surfaces, were prepared for targeted chemotherapy synergized with trans-spatial immunotherapy. Initially, these nanoparticles could selectively deliver DOX into HSP70-overexpressing tumor cells for targeted chemotherapy and thus antigen production. Next, these nanoparticles could be biodegraded into TKD and CD-involved debris, which therefore could largely adsorb the tumor antigens, probably via π-π conjugation synergized hydrophobic interactions, and then adequately transport these antigens from the immunosuppressive TME into CD91-overexpressing DCs in the spleen due to the special TKD-mediated small size effects. Finally, owing to TKD-mediated lysosome endocytosis, these antigens carried by the debris could be processed into antigen peptides, which thus could be combined with MHC to construct the MHC complex and then present them on DC cytomembranes to activate T cells. Consequently, enhanced immunotherapy with efficient antigen presentation and avoided TME immunosuppression was accomplished, synergized with targeted chemotherapy, remarkably improving the therapeutic outcomes of even triple-negative breast cancer.

Declaration of competing interest

The authors declare that they have no conflicts of interest in this work.

Biographies

Yi Wang received his B.S. in 2004 from Wuhan University, M.S. in 2007 and Ph.D. in 2014 from Fudan university. He is now a young professor in Center for Advanced Low-dimension Materials, Donghua University. His current interests are the design and synthesis of advanced materials for biomedical and energy applications.

Rongqin Huang(BRID: 09685.00.85520) is now a full professor in School of Pharmacy, Fudan University. She obtained her B.S. in 2003 and Ph.D. in 2008 from Fudan University. She finished her researches in Munich University in 2010 and UC Berkeley from 2015 to 2016. Her research interests include targeted drug delivery, combined cancer therapy and biomaterials.