Biological chemotaxis-guided self-thermophoretic nanoplatform augments colorectal cancer therapy through autonomous mucus penetration

Abstract

Oral drug delivery systems have great potential to treat colorectal cancer (CRC). However, the drug delivery efficiency is restricted by limited CRC-related intestine positioning and dense mucus barrier. Here, we present a biological chemotaxis-guided self-thermophoretic nanoplatform that facilitates precise intestinal positioning and autonomous mucus penetration. The nanoplatform introduces asymmetric platinum-sprayed mesoporous silica to achieve autonomous movement in intestinal mucus. Furthermore, inspired by the intense interaction between pathogenic microbes and CRC, the nanoplatform is camouflaged by Staphylococcus aureus membrane to precisely anchor in CRC-related intestine. Owing to 4.3-fold higher biological chemotactic anchoring of CRC-related intestine and 14.6-fold higher autonomous mucus penetration performance, the nanoplatform vastly improves the oral bioavailability of cisplatin, leading to a tumor inhibition rate of 99.1% on orthotopic CRC–bearing mice. Together, the exquisitely designed nanoplatform to overcome multiple physiological barriers provides a new horizon for the development of oral drug delivery systems.

The nanoplatform with biological chemotaxis and mucus penetration abilities provides a feasible strategy for oral drug delivery.

INTRODUCTION

Colorectal cancer (CRC) is one of the leading causes of cancer incidence and mortality worldwide, which seriously threatens the life of patients (1, 2). Oral administration is a convenient and simple administration route to treat CRC and other digestive tract cancers (3, 4). With the development of nanotechnology, drug delivery systems (DDSs) provide advantages for oral treatment of CRC due to their function of protecting drugs from the gastrointestinal environment, reducing toxicity, and improving efficiency (5, 6). However, the dense and viscous structure of intestinal mucus in the colorectum severely limited the penetration efficiency of DDS into CRC sites. Intestinal mucus is a notable biological barrier to protect intestinal epithelial cells from the invasion of complex intestinal environments and microorganisms. Nevertheless, the intestinal mucus barrier seriously hindered the delivery efficiency of oral drugs, which is the bottleneck of effective drug delivery for CRC therapy (7, 8). How to efficiently penetrate intestinal mucus is the key to improving the bioavailability of chemotherapeutic drugs and the efficiency of CRC therapy.

In recent years, the hydrophilic modified DDS (9) and mucinase (such as bromelain)–modified DDS (10) were developed to penetrate the intestinal mucus barrier. For example, He et al. (11) used polyethylene glycol (PEG) to modify nanoplatform with penetration efficiency of ~2.8%. However, the active penetration ability and penetration depth of the above methods are still limited (12). Therefore, it is urgent to develop an efficient and nondestructive intestinal mucus penetration strategy to improve the bioavailability of DDS. Recently, nanomotors have gradually attracted the attention because of the superior and controllable motion ability (13, 14). Among them, near-infrared (NIR) laser–propelled nanomotors has broad application prospects on complex physiological barrier penetration because of the characteristics of fuel-free and stable motion performance (15, 16). Ji et al. (17) reported that the penetration efficiency of nanomotors to the stratum corneum of the skin was 2.5 times higher than that of passive diffusion nanoparticles. The remarkable motility characteristic of nanomotors offers great opportunities for the penetration of intestinal mucus barrier. Nevertheless, there was no previous literature about the penetration of nanomotors to the intestinal mucus. Before penetrating intestinal mucus for the treatment of CRC, it remains a severe challenge on DDS to selectively position the intestinal tract of CRC, which hinders the targeted enrichment and intestinal penetration of DDS to CRC. Because of the long and complex intestinal structure, it is particularly important for DDS to accurately locate the intestinal segment of CRC.

Bionics provides opportunities for nanotechnology and targeted drug delivery (18, 19). The activation of oncogenes in CRC cells leads to abnormal activation of inflammation-related transcription factors (20). Hence, the occurrence and development of CRC are accompanied by the emergence of intestinal inflammation. The inflammatory environment of CRC secretes a large amount of immunoglobulin G (IgG). Furthermore, the inflammatory environment of CRC has intensive interaction with pathogenic microorganisms. Staphylococcus aureus can specifically bind to IgG in the inflammatory environment of CRC by protein A on the bacterial surface. Therefore, S. aureus can sensitively sense the inflammatory environment, specifically recognize IgG, and efficiently colonize the intestinal tract of CRC (21, 22). However, the direct application of S. aureus in the delivery system has certain pathogenic risks (23). Our previous work showed that bacterial biomimetic membrane camouflage could provide excellent gut-oriented delivery performance (24). Extraction of S. aureus biomimetic membrane (SAM) and encapsulation of nanomotors also retained the inflammatory homing and intestinal colonization effects of IgG with superior biological safety.

Here, to improve the oral drug delivery efficiency, we present a biological chemotaxis-guided self-thermophoretic nanoplatform (BCTN)-augmented CRC therapy through precise intestinal positioning and autonomous mucus penetration (Fig. 1). The nanoplatform introduces high drug-loading mesoporous silicon as matrix and is partially sprayed with platinum to construct asymmetric directional autonomous moveable nanoplatform. Subsequently, SAM is used to camouflage the nanoplatform because of the intensive interaction between intestinal pathogenic microbes and the inflammatory environment of CRC and provides a prerequisite for moveable nanoplatform to actively penetrate the intestinal microenvironment of CRC. The camouflage of SAM can specifically target the intestinal inflammatory environment of CRC. After substantial anchoring to intestinal sites of CRC, BCTN asymmetrically absorbs the NIR laser and autonomously penetrates the dense and viscous mucus barrier by self-thermophoretic propulsion force. In addition, BCTN can rapidly cross the intestinal epithelium and is massively taken up by CRC cells via the glucose transporter (GLUT) pathway. Furthermore, high concentration of reduced glutathione (GSH) triggers the rapid release of chemotherapeutic drug cisplatin (CP) and induces apoptosis of CRC cells. With accurate tumor microenvironment recognition and efficient intestinal mucus penetration abilities, BCTN can produce a tumor inhibition rate of 99.1% on orthotopic CRC tumor–bearing mice, which provides a new horizon for the development of oral DDSs.

Fig. 1. A schematic of the synthetic procedure and CRC therapy of BCTN.

RESULTS

Preparation and characterization of BCTN

The main preparation procedure and delivery mechanism of BCTN were illustrated in Fig. 1. In this work, mesoporous silica nanoparticles (MSNs) were selected as the carrier of the nanoplatform because of its large drug-loading capacity and excellent biocompatibility (25, 26). Then, platinum was unsymmetrically sputtered on the surface of MSN to prepare asymmetric structure (SiO₂/Pt). The transmission electron microscope (TEM) image reveals that the ~8-nm thickness of platinum layer was coated on the surface of SiO₂/Pt (Fig. 2A). Moreover, elemental mapping images in Fig. 2B proved the asymmetric spray plating of platinum. Furthermore, the elemental analysis identified that the platinum content at the sputtered side was 38.1%, which was only 0.6% at the MSN side (Fig. 2C). To improve the physicochemical properties of nanoplatform, SiO₂/Pt was functionalized with SSPEG₂₀₀₀-Glu (PEG-Glu) at the disulfide bond site. After functionalization of PEG-Glu, the particle size increased from 48.5 ± 2.6 to 51.4 ± 5.2 nm, and the zeta potential changed from −21.9 ± 0.6 to 3.0 ± 1.3 mV (Fig. 2, D and E). Benefitting from the large drug-loading capacity of MSN, the drug-loading content of the nanoplatform for CP was 21.2 ± 0.9%, with the encapsulation efficiency of 28.7 ± 1.3%. Furthermore, thermogravimetric analysis confirmed the component content of BCTN (Fig. 2F and fig. S1).

Fig. 2. The preparation and characterization of BCTN.

(A) TEM images of MSN, SiO₂/Pt, and SiO₂/Pt@PEG-Glu and negative staining TEM images of SAM and BCTN. (B) Element mapping image of SiO₂/Pt@PEG-Glu@SAM. (C) Elemental analysis of Pt in SiO₂/Pt@PEG-Glu@SAM. Particle size (D) and zeta potentials (E) of SiO₂, SiO₂/Pt, SiO₂/Pt@PEG-Glu, and BCTN (n = 3). (F) Component content of BCTN tested by thermogravimetric analysis (n = 3). (G) Energy-dispersive x-ray spectrum of SiO₂/Pt@PEG-Glu@SAM. (H) Western blot analysis of protein A in S. aureus, SAM, and BCTN. GAPDH, glyceraldehyde-3-phosphate dehydrogenase. (I) Content of enterotoxins in S. aureus, SAM, and BCTN (n = 3). Values are represented as means ± SD. Significance was calculated via one-way analysis of variance (ANOVA) with Tukey’s posttest. **P < 0.01.

Owing to the intensive interaction between specific microorganisms and the inflammatory environment of CRC, SAM was used to camouflage the nanoplatform to achieve biological chemotaxis to inflamed intestinal sites of CRC. After the SAM coating, BCTN (about 80 nm) had a ~10-nm shell according to the negative staining TEM image in Fig. 2A. Furthermore, a broader view of TEM results of BCTN in fig. S2 confirmed the BCTN with uniform particle size distribution. We also detected the polydispersity index value (0.124 ± 0.02) of BCTN by dynamic light scattering, indicating the superior uniformity of BCTN. The particle size and zeta potential of SAM were 411.0 ± 115.8 nm and −19.8 ± 0.9 mV, respectively (fig. S3). Because of the SAM coating on the surface, BCTN showed a significantly increased particle size of 81.1 ± 4.9 nm versus 51.4 ± 5.2 nm for the SiO₂/Pt@PEG-Glu, and a decreased zeta potential of −8.3 ± 1.0 mV versus 3.0 ± 1.3 mV for the SiO₂/Pt@PEG-Glu (Fig. 2E). In addition, the element analysis in Fig. 2G confirmed the successful camouflage of SAM. Furthermore, the characteristic peak in Fourier transform infrared spectroscopy verified the existence of the biological structure of SAM (fig. S4). Moreover, the fluorescence colocalization experiment of SiO₂/Pt@PEG-Glu and SAM also confirmed the camouflage of SAM (fig. S5). In addition, Western blot analysis showed that the extraction and coating process had no impact on protein A on the biomimetic membrane (Fig. 2H). Enterotoxin is one of the main factors why S. aureus causes intestinal inflammation. After the extraction and coating process, the content of enterotoxin in nanoplatform significantly decreased (Fig. 2I). Therefore, compared with S. aureus as direct drug carrier, the biosafety of nanoplatform after SAM camouflage was improved considerably.

BCTN holds superior biological chemotaxis and autonomous movement characteristics

Efficient and precise anchoring of the nanoplatform in the intestine of CRC is the key that affects the therapeutic efficacy of CRC. There is an inflammatory environment in the intestinal tract of CRC, and a large amount of Ig is secreted by the intestinal immune system (27). To verify the recognition ability of SAM-coated nanoparticles to IgG, we first examined the adsorption of nanoparticles to IgG-labeled magnetic beads (Fig. 3A). The adsorption experiment results showed that BCTN after SAM coating had a better adsorption effect on magnetic beads compared with other nanoparticles (Fig. 3B). BCTN could be rapidly adsorbed on IgG-labeled magnetic beads within 10 min, laying solid foundation to the rapid recognition and anchoring in IgG-rich CRC intestinal segments (fig. S6). Besides, the adsorption effect disappeared when BCTN was preincubated with IgG. Together, the construction of biological chemotactic self-thermophoretic nanoplatform by encapsulation of SAM helps exert IgG anchoring and biological-chemotactic effects in the intestinal inflammatory region of CRC.

Fig. 3. BCTN holds superior biological chemotaxis and autonomous movement characteristics.

(A) Schematic diagram of the adsorption of nanoparticles to IgG-labeled magnetic beads. (B) Fluorescence intensity of magnetic beads after incubation with different nanoparticles (n = 3). a.u., arbitrary units. Representative tracking trajectory (C) and motion speed (D) of nanoparticles in mucin solution with varied laser power over 10 s (0, 0.5, 1, and 1.5 W cm⁻²; n = 10). MSD of SiO₂/Pt (E) and BCTN (F) in mucin solution under different power densities of lasers (n = 10). (G) Colocalization rate of SiO₂/Pt@PEG-Glu and SAM after being dispersed in different physiological environment simulation solutions (n = 3). (H) Drug release rate of CP@SiO₂/Pt, CP@SiO₂/Pt@PEG-Glu, and BCTN in phosphate-buffered saline (PBS; pH 7.4) and GSH containing PBS (pH 7.4, n = 3). Values are represented as means ± SD. Significance was calculated via one-way ANOVA with Tukey’s posttest. **P < 0.01 and ***P < 0.001.

Next, the motion performance of the nanoplatform in intestinal mucus was verified. The intestinal mucus in the human colorectum has a dense and viscous structure with a thickness of 200 to 800 μm (28). Therefore, the movement of DDS in the intestinal mucus barrier is seriously restricted, resulting in limited penetration ability and bioavailability. By combining nanomotors with high spatiotemporally resolved NIR laser, BCTN could asymmetrically absorb NIR laser to produce the self-thermophoretic effect. The representative tracking trajectories of MSN, SiO₂/Pt, and BCTN in mucin solution under different power densities of NIR lasers were shown in Fig. 3C. Without Pt spraying, MSN still produced a slow movement in mucin solution after 808-nm NIR laser irradiation (1.5 W cm⁻²; movie S1). In sharp contrast, the motion speed of SiO₂/Pt in mucin solution remarkably accelerated and reached 2.1 ± 0.2 μm s⁻¹ (movie S2). The motion speed of BCTN was 2.0 ± 0.2 μm s⁻¹ after NIR laser irradiation (Fig. 3D and movie S3), which had no significant difference compared with that of SiO₂/Pt. In addition, with the increase of laser power, the motion speed and mean square displacement (MSD) of the nanoplatform also significantly increased (Fig. 3, E and F). The advanced motion speed inspired the nanoplatform to penetrate the intestinal mucus barrier autonomously and vastly improve the delivery efficiency.

To explore the protective effect and shedding behavior of SAM from the nanoplatform, we dispersed BCTN in different physiological simulation fluids for 2 hours (Fig. 3G and fig. S7). After 2 hours in artificial gastric fluid, most of the SAM did not shed from the nanoplatform, indicating that the bacterial biomimetic membrane was able to resist gastric acid digestion. Similarly, SAM hardly shed after being dispersed in artificial intestinal fluid. Negative staining TEM image of BCTN after incubation in artificial gastric fluid and intestinal fluid for 2 hours also confirmed outer membrane integrity of BCTN (fig. S8), whereas SAM gradually shed after dispersing in the artificial colon fluid (fig. S9). After reaching the colon, 30% of BCTN was still coated with SAM after 2 hours. When incubated in artificial colon fluid for 8 hours, 14.8 ± 4.4% of SAM was coated on BCTN (fig. S10). The results indicated that SAM would naturally shed from the surface of the nanoplatform after long incubation time in artificial colon fluid. Besides, the colorectal microenvironment harbors a large variety of bacteria, and the bacteria release many substances such as bacteriocins and ferritins (29, 30), resulting in degradation of the bacterial biomimetic membrane. Therefore, the bacterial biomimetic membrane could shed under the specific stimulation of the colorectal environment, exposing the PEG-Glu structure to facilitate the affinity on intestinal endothelial cells.

To evaluate the in vitro drug release behavior before and after SAM coating, CP@SiO₂/Pt@PEG-Glu and BCTN were dispersed in different physiological solutions (Fig. 3H). According to the release results, the drug release rate of BCTN in phosphate-buffered saline (PBS) at 2 and 4 hours were 14.2 ± 2.5% and 22.0 ± 2.3%, respectively, which was significantly lower than CP@SiO₂/Pt (23.2 ± 1.4% and 38.3 ± 1.6%, respectively) because of the PEG-Glu structure, and most of the drug was still encapsulated in the nanoplatform. The drug release from BCTN in artificial colonic fluid slightly improved compared with that in PBS because of the gradual shedding of SAM after incubation in artificial colonic fluid (fig. S11). Nevertheless, the drug release rate of BCTN at 2 and 4 hours in artificial colonic fluid was only 15.3 ± 0.9% and 24.5 ± 0.9%. Therefore, the relative slow drug release also laid foundation for the nanoplatform carrying majority of CP across the intestinal mucus barrier after NIR laser irradiation.

Nevertheless, the drug release efficiency of CP@SiO₂/Pt@PEG-Glu was significantly accelerated when dispersed in GSH containing PBS (pH 7.4). The release rate of CP from CP@SiO₂/Pt@PEG-Glu in GSH containing PBS after 12 hours was 57.8 ± 1.9%, which was close to that of CP@SiO₂/Pt (61.8 ± 1.7%). Because of the existence of S─S bond in the PEG-Glu structure, GSH would break the S─S bond, resulting in the shed of PEG-Glu and rapid release of CP. Moreover, comparing the drug release curve before and after NIR laser irradiation (1.5 W cm⁻²), it could be concluded that NIR laser irradiation had no significant effect on the drug release behavior of nanoplatform, which laid the foundation for the nanoplatform to efficiently penetrate the intestinal barrier in the intestine.

BCTN efficiently penetrates intestinal mucus barrier

The above experiments demonstrated the superior motion property of the nanoplatform. To verify the penetration performance of the nanoplatform across the intestinal mucus barrier, we incubated the nanoplatform with rat colorectum and used high spatiotemporally resolved NIR laser for fixed-point irradiation and detected the penetration depth of the nanoplatform on the intestinal mucus barrier by confocal laser scanning microscope (CLSM). The CLSM images in Fig. 4A showed that only SiO₂/Pt@PEG-Glu after NIR laser irradiation (1.5 W cm⁻²) could actively penetrate the intestinal mucus barrier, which meant that asymmetric spraying of platinum and NIR laser irradiation were necessary for autonomous penetration of the intestinal mucus barrier. To quantify the mucus penetrability of nanoplatform, we estimated that the thickness of rat intestinal mucus was 54.4 ± 6.2 μm (Fig. 4B). After 5 min of laser irradiation, the penetration depth of SiO₂@PEG-Glu was only 15.2 ± 4.4 μm. In sharp contrast, the penetration depth of SiO₂/Pt@PEG-Glu was 67.9 ± 8.7 μm, which indicated that SiO₂/Pt@PEG-Glu could effectively penetrate the intestinal mucus barrier after NIR laser irradiation.

Fig. 4. BCTN efficiently penetrates intestinal mucus barrier.

(A) CLSM fluorescent images of penetration depth of intestinal mucus barrier through different nanoparticles. (B) Mean penetration depth measured in (A) (n = 5). Mean fluorescence intensity (MFI) measured in Caco-2 cells (C) and CT26 cells (D) by CLSM at 3 hours after incubation with different nanoparticles (n = 3). (E) Schematic diagram of mucus-producing transwell model composed of Caco-2 and HT29-MTX cells. (F) CLSM images of CT26 cells in basal chamber after 3 hours of incubation with different nanoparticles in apical chamber. (G) MFI measured in CT26 cells of basal chamber after 3 hours of incubation with different nanoparticles in apical chamber (n = 3). (H) Average number of nanoparticles in TEM images of intestinal loop and villi (n = 3). (I) MFI of nanoparticles in intestinal submucosa (n = 3). (J) TEM images of distribution of different nanoparticles in the intestinal loop and villi. (K) CLSM images of colocalization of cell nucleus and mucus with different nanoparticles in colorectum after 5 min of incubation with nanoparticles; SiO₂/Pt@PEG-Glu + NIR group received NIR laser irradiation (1.5 W cm⁻²) for 5 min during the incubation. Values are represented as means ± SD. Significance was calculated via one-way ANOVA with Tukey’s posttest. *P < 0.05, **P < 0.01, and ***P < 0.001. DAPI, 4′,6-diamidino-2-phenylindole.

After verifying the excellent intestinal mucus penetration ability of the nanoplatform, the enterocyte transcytosis and tumor cell uptake ability of the nanoplatform were next detected. As mentioned above, the intestinal barrier includes not only the intestinal mucus barrier but also the enterocyte barrier. As a critical cell for nutrient absorption, enterocyte is selective for the absorption of various substances. Therefore, the rapid transcytosis of enterocyte is conducive to improving the efficiency of drug delivery and bioavailability. As one of the eight essential nutrients for the human body, glucose rapidly enters enterocyte via the intestinal Na⁺-dependent glucose transporter and enters the blood and intestinal tissues via facilitated diffusion mediated by glucose transporter 2 (31).

Caco-2 cells, human colonic adenocarcinoma cells, are widely used as the enterocyte monolayer models to predict the cellular uptake behavior of nanoplatform (32). After Caco-2 cells were incubated with different nanoparticles for 3 hours, and the cellular uptake of nanoparticles was measured by the semiquantitative analysis of fluorescence intensity in Caco-2 cells. After modification with PEG-Glu, SiO₂/Pt@PEG-Glu exhibited more cellular uptake on Caco-2 monolayer cell models (Fig. 4C and fig. S12). The mean fluorescence intensity (MFI) of the SiO₂/Pt@PEG-Glu group was 1.6-fold higher than that of the SiO₂/Pt@PEG group. To explore the cellular uptake pathway of nanoplatform, dapagliflozin was used to inhibit the glucose transporter on the Caco-2 cell membrane (33). After inhibiting glucose transporter on the Caco-2 cell membrane, the uptake of nanoparticles by Caco-2 cells significantly reduced, which indicated that the cellular uptake of the nanoparticles required the participation of glucose transporters. The MFI of the SiO₂/Pt@PEG-Glu after NIR laser irradiation was twofold higher than that of the SiO₂/Pt@PEG-Glu. In addition, the results of flow cytometry analysis showed that the SiO₂/Pt@PEG-Glu + NIR group induced more cellular uptake in Caco-2 cells compared with that of other groups (fig. S13). The increased internalization speed of Caco-2 cells may be mainly due to the remarkably accelerated motion speed of nanoplatform after NIR laser irradiation. In addition, thermostimulation may increase the uptake of cancer cells (34).

Since malignant cells rely on the Warburg effect for energy production, the glucose transport efficiency is significantly higher in malignant cancer cells than that in normal cells, which depends on the increased expression of glucose transporters (35, 36). Therefore, the nanoplatform was continued to investigate the cellular uptake performance of CT26 cancer cells (Fig. 4D and fig. S14). Because of the surface modification of Glu, the MFI of SiO₂/Pt@PEG-Glu in CT26 cells was 1.7-fold higher than that of SiO₂/Pt@PEG. After incubation with dapagliflozin, the uptake of nanoparticles by CT26 cells was significantly reduced, which indicated that the cellular uptake of nanoparticles also required the participation of glucose transporters. After NIR laser irradiation for 5 min, MFI of the SiO₂/Pt@PEG-Glu + NIR group was 1.4-fold higher than that of the SiO₂/Pt@PEG-Glu group, which indicated that NIR laser irradiation could increase the uptake of CT26 cells.

Encouraged by the superior cellular uptake ability on Caco-2 and CT26 monolayer cell models, the nanoplatform examined the penetration performance across the mucus-producing intestinal barrier. An in vitro mucus-producing transwell model was constructed with mixed Caco-2 and HT29-MTX cells (the ratio of cells was 7:3) in the apical chamber of the transwell plate (Fig. 4E). The CLSM images in Fig. 4F revealed that after PEG-Glu modification, SiO₂/Pt@PEG-Glu could effectively penetrate the in vitro mucus-producing intestinal barrier and be taken up by CT26 cells in the basal chamber of transwell plate (P < 0.05 compared with SiO₂/Pt@PEG and P < 0.01 compared with SiO₂/Pt). Furthermore, after 5 min of NIR laser irradiation (1.5 W cm⁻²), MFI of the SiO₂/Pt@PEG-Glu + NIR group was 2.9- and 2.1-fold higher than that of the SiO₂/Pt group and SiO₂/Pt@PEG-Glu group, respectively (Fig. 4G). The results implied that SiO₂/Pt@PEG-Glu showed enhanced mucus penetration and accelerated endocytosis by mixed Caco-2 cells and HT29-MTX cells after NIR laser irradiation. In addition, flow cytometry analysis also confirmed that considerable nanoparticles were internalized by CT26 cells in the basal chamber in the SiO₂/Pt@PEG-Glu + NIR group (fig. S15).

In view of the aforementioned marked in vitro penetration properties, the nanoplatform was continued to examine the in vivo penetration performance on intestinal mucus barrier of CRC tumor–bearing mice. CT26 transfected with luciferase (CT26-luc) cells were injected into the cecal wall of the mice to establish an orthotopic CRC tumor–bearing mouse model as previously described with slight modification (37). To evaluate the intestinal absorption and tumor penetration performance, the in situ intestinal loop model was used (38). After circulation perfusion with the nanoplatform for 5 min, the intestines of tumor site were taken and measured by TEM and CLSM. The TEM images of intestinal loop and villi in SiO₂/Pt@PEG-Glu group showed that the average number of nanoparticles reaching the enterocyte was moderately increased compared with SiO₂/Pt and SiO₂/Pt@PEG groups. Owing to the functionalization of PEG-Glu, the average number of nanoparticles reaching the enterocyte in SiO₂/Pt@PEG-Glu group was 3.0- and 2.6-fold higher than that of the SiO₂/Pt and SiO₂/Pt@PEG groups, respectively (Fig. 4, H and J). Furthermore, after NIR laser irradiation (1.5 W cm⁻²) for 5 min, SiO₂/Pt@PEG-Glu could effectively penetrate the intestinal mucus barrier to reach the enterocyte. The average number of nanoparticles reaching the enterocyte in the SiO₂/Pt@PEG-Glu + NIR group was twofold higher than that of the SiO₂/Pt@PEG-Glu group (P < 0.001). The results showed that more nanoparticles crossed the intestinal mucus barrier in the SiO₂/Pt@PEG-Glu + NIR group compared with other groups, which increased affinity of the nanoparticles to intestinal epithelial cells.

After penetrating the intestinal mucus barrier, the nanoparticles contacted with the intestinal epithelial cells and may be transcytosed by the intestinal epithelial cells to reach intestinal submucosa of the colorectum. Therefore, the accumulation of nanoparticles in the intestinal submucosa was further studied by CLSM. After 5 min of NIR laser irradiation (1.5 W cm⁻²), SiO₂/Pt@PEG-Glu rapidly penetrated the intestinal mucus barrier and the enterocyte barrier, resulting in a significant increase of the fluorescence intensity in the intestinal submucosa (Fig. 4, I and K). The MFI of intestinal submucosa in the SiO₂/Pt@PEG-Glu + NIR group was 128.2 ± 42.9, which was 14.6-, 2.8-, and 1.9-fold higher than that of the SiO₂/Pt, SiO₂/Pt@PEG, and SiO₂/Pt@PEG-Glu groups, respectively, which proved the superior in vivo multiple intestinal barrier penetration ability of nanoplatform after NIR laser irradiation. The improved accumulation capacity in intestinal submucosa was mainly due to the self-thermophoretic propulsion of SiO₂/Pt@PEG-Glu after NIR laser irradiation. In addition, the functional modification of PEG-Glu also improved the accumulation of nanoparticles in intestinal submucosa. After incubation with the nanoplatform for 5 min, the MFI of intestinal submucosa in the SiO₂/Pt@PEG-Glu group was 68.6 ± 6.6, which was 7.8-fold and 1.5-fold higher than that of the SiO₂/Pt and SiO₂/Pt@PEG groups, respectively. On the basis of the in vitro and in vivo multiple intestinal barrier penetration experiments, BCTN could efficiently penetrate the intestinal barrier, which may provide a new horizon for the crossing of multiple complex physiological barriers.

BCTN exhibits efficient in vitro anticancer activity

After verifying the excellent intestinal barrier penetration performance of BCTN, in vitro anticancer performance of the drug-loading nanoplatform was next examined. The in vitro biosafety of the nanoplatform was first evaluated. The empty nanoplatform was incubated with Caco-2 and CT26 cells for 24 hours, and the results indicated that even the concentration of the nanoplatform reached 200 μg ml⁻¹, the cell viability of Caco-2 and CT26 cells was still more than 90%, indicating excellent in vitro biosafety of the nanoplatform (fig. S16, A and B). We also conducted in vitro cytotoxic experiments of NIR laser irradiation alone. The cell viability of Caco-2 and CT26 cells treated with different power densities of NIR lasers still kept more than 90%, indicating the negligible toxicity of NIR laser irradiation (fig. S16, C and D). Afterward, the drug-loading nanoplatform was incubated with CT26 cells for 24 hours, and the cell viability of CT26 cells was shown in Fig. 5A. When the drug concentration was 5 μg ml⁻¹, the cell viability of CT26 cells in the CP@SiO₂/Pt@PEG-Glu with NIR laser irradiation (1.5 W cm⁻²) group was only 16.6 ± 0.7%, with highly significant statistical difference compared with the CP, CP@SiO₂/Pt, and CP@SiO₂/Pt@PEG-Glu groups (45.5 ± 0.7%, 58.2 ± 0.8%, and 46.3 ± 2.8%, respectively). The cell viability of HT29 cells (human colon carcinoma cell lines) after incubation with the drug-loading nanoplatform was similar to that of CT26 cells (Fig. 5B). When the drug concentration reached 5 μg ml⁻¹, the cell viability of HT29 cells in the CP@SiO₂/Pt@PEG-Glu + NIR group was only 18.5 ± 1.0%, with highly significant statistical difference compared with the CP, CP@SiO₂/Pt, and CP@SiO₂/Pt@PEG-Glu groups (72.7 ± 1.3%, 76.6 ± 1.0%, and 60.3 ± 1.9%, respectively).

Fig. 5. BCTN exhibits efficient in vitro anticancer activity.

Cell viability of CT26 (A) and HT29 (B) cancer cells after incubation with different nanoplatforms for 24 hours (n = 3). (C) Ratio of dead CT26 cells after incubation with different nanoplatforms for 24 hours (5 μg ml⁻¹, in terms of CP; n = 3). (D) Live/dead cell staining fluorescence images of CT26 cells stained by calcein AM/propidium iodide (PI) after incubation with different nanoplatforms for 24 hours (scale bars, 100 μm). (E) Schematic diagram of mucus-producing transwell model composed of Caco-2 and HT29-MTX cells to examine anticancer activity of CP and different nanoplatforms across the apical chamber. (F) Flow cytometric analysis of CT26 cells in mucus-producing transwell model after incubation with the nanoplatform for 24 hours. Values are represented as means ± SD. Significance was calculated via one-way ANOVA with Tukey’s posttest. ***P < 0.001. FITC, fluorescein isothiocyanate.

To intuitively observe the ratio of live cells and dead cells, CT26 cells were incubated with drug-loading nanoplatform to conduct live/dead (green/red) staining experiment. The proportion of living and dead cells in each group was quantified and presented in Fig. 5C. Compared with the intense green fluorescence in other groups, the green fluorescence sharply decreased in the CP@SiO₂/Pt@PEG-Glu + NIR group (Fig. 5D), which was significantly lower than that in other groups (P < 0.001). The percentage of dead cells in the CP@SiO₂/Pt@PEG-Glu + NIR group was 78.6 ± 0.5%, which proved the excellent in vitro anticancer activity of the CP@SiO₂/Pt@PEG-Glu after NIR laser irradiation (1.5 W cm⁻²).

To investigate whether the drug-loading nanoplatform could produce excellent anticancer activity after penetrating the intestinal barrier, the in vitro mucus-producing transwell model was applied to evaluate the cancer cell viability in the basal chamber after the incubation of the drug-loading nanoplatform with mixed Caco-2 and HT29 cells in the apical chamber (Fig. 5E). Afterward, the CT26 cells in the basal chamber were analyzed by flow cytometry, and the results were presented in Fig. 5F. The percentage of apoptotic and necrotic cells in the CP@SiO₂/Pt@PEG-Glu + NIR group was 46.7%, much higher than the 14.5% in the CP@SiO₂/Pt group and 20.1% in the CP@SiO₂/Pt@PEG-Glu group, respectively. The promising intestinal barrier penetration ability of nanoplatform in Fig. 4F and accelerated cancer cell uptake of nanoplatform in Fig. 4D after NIR laser irradiation may contribute to the facilitated in vitro anticancer activity. In addition, it was reported that platinum nanoparticles would degrade tumor-endogenous H₂O₂ to oxygen after laser irradiation, and the generated oxygen might sensitize the chemotherapeutic effect of CP (39, 40). The above results indicated that the drug-loading nanoplatform not only had the ability to autonomously penetrate the intestinal barrier but also showed excellent apoptosis-inducing effect on CRC cells after crossing the intestinal barrier.

BCTN precisely anchors in the intestinal tract of CRC

Before the in vivo distribution experiments, the in vivo biosafety of the nanoplatform was examined. During the 3 weeks of oral administration of nanoplatform, the body weight change of mice had no significant difference between the BCTN group and control group (fig. S17). At the end of oral administration, the mice in each group were sacrificed to take main tissues for hematoxylin and eosin (H&E) staining to check the changes of tissue physiology (fig. S18). The results indicated that oral administration of BCTN and NIR laser irradiation did not cause significant organ damage. The representative inflammation cytokines such as tumor necrosis factor–α (TNF-α) and interleukin-6 (IL-6) in the blood of mice in each group were detected. The TNF-α and IL-6 in both BCTN and BCTN + NIR–treated mice had no significant difference compared with control group (fig. S19), indicating that the nanoplatform did not cause significant inflammation in mice. In addition, we detected the representative enzymes in serum biochemical indexes, such as alanine aminotransferase (ALT) and aspartate aminotransferase (AST), and the results were as shown in fig. S20. The ALT and AST of mice in BCTN and BCTN + NIR groups did not have significant change compared with the control group. In addition, other serum biochemical indexes such as total bilirubin, total protein, blood urea nitrogen, and serum creatinine A indicated that BCTN and NIR laser irradiation did not cause significant liver and kidney damage. Besides, the hematological parameters of mice showed that the mice in NIR group, BCTN group, and BCTN + NIR group had negligible changes compared with control groups (fig. S21). These results indicated that the nanoplatform had no significant toxicity during the 3 weeks of treatment.

Motivated by the above excellent intestinal barrier penetration ability and superior in vivo biosafety, the in vivo biodistribution of the nanoplatform was investigated in an orthotopic colorectal tumor–bearing mice model, which was confirmed by H&E staining for the successful establishment (fig. S22). Efficient and accurate anchoring of the DDS in the intestinal tract of CRC is the priority in exerting therapeutic effects on CRC. Therefore, three common bacteria in the intestine were selected to evaluate the targeting performance to CRC-related intestine after the camouflage of nanoplatform (Fig. 6A). Three kinds of bacterial membrane–coated nanoparticles (SAM-coated SiO₂/Pt@PEG-Glu as BCTN, bifidobacteria membrane–coated SiO₂/Pt@PEG-Glu as SiO₂/Pt@PEG-Glu@BFM, and Escherichia coli membrane–coated SiO₂/Pt@PEG-Glu as SiO₂/Pt@PEG-Glu@ECM) were prepared with approximately equivalent bacterial membrane amounts (fig. S23). It can be intuitively seen that the intestinal retention effect of the nanoplatform was significantly improved after bacterial biomimetic membrane coating (fig. S24), which was consistent with our previous study (24), but the targeting ability of the nanoplatform to CRC-related intestine varied because of different bacterial biomimetic membrane camouflage. The nanoplatform after the camouflage of BFM and ECM lack recognition of CRC-related intestine. The fluorescence at the CRC-related intestine was not significantly different from that of the uncoated nanoplatform (Fig. 6B). However, the accumulation properties of BCTN in CRC-related intestine were enhanced considerably after SAM coating compared with that of other groups (P < 0.01).

Fig. 6. BCTN precisely anchors in the intestinal tract of CRC.

(A) Distribution fluorescence images of gastrointestinal tract and main organs in tumor-bearing mice after oral administration of different nanoparticles for 12 hours (the red arrow points to the location of the tumor; H, heart; Li, liver; S, spleen; Lu, lung; K, kidney). (B) MFI of intestinal segment of CRC and main organs in tumor-bearing mice after oral administration of different nanoparticles (n = 3). (C) Distribution fluorescence images of intestine and main organs in tumor-bearing mice after oral administration of nanoparticles at different time points. (D) MFI of intestinal segment of CRC in tumor-bearing mice after oral administration of nanoparticles at different time points (n = 3). (E) MFI of major organs of tumor-bearing mice after oral administration of different nanoparticles for 24 hours (n = 3). (F) CLSM images of the intestine sections of CRC at 5 min after incubation with different nanoparticles (the white line indicates the location of the tumor sites). (G) MFI of different nanoparticles in the tumor site (n = 3). (H) Plasma concentration time curve after oral administration of different nanoparticles for 24 hours (n = 3). Values are represented as means ± SD. Significance was calculated via one-way ANOVA with Tukey’s posttest (B, D, E, and G) and two-way ANOVA with Dunnett’s posttest (H). **P < 0.01 and ***P < 0.001.

To assess the distribution of nanoplatform at different gastrointestinal tracts, we investigated the SAM-mediated intestinal localization of nanoplatform by immunofluorescence images (fig. S25). The semiquantitative analysis was shown in fig. S26. The MFI in colon after oral administration of BCTN was significantly higher than that of SiO₂/Pt@PEG-Glu, demonstrating the superior anchoring ability to intestinal segment of CRC.

Having demonstrated the accurate anchoring of BCTN to CRC-related intestine, we next evaluated the targeting effect of the nanoplatform at different times. After the mice received oral administration of the nanoplatform, the fluorescence of mice in each group was recorded by small animal imaging system at time intervals (Fig. 6C). The locations of the tumor sites were marked by red circles according to the bioluminescence images of the tumor sites in the gastrointestinal tract (fig. S27). BCTN showed a stronger and more prolonged fluorescence in the tumor site of orthotopic colorectal tumor–bearing mice compared with uncoated nanoparticles. Quantitative analysis of the fluorescence at intestinal segment of CRC was shown in Fig. 6D. The fluorescence of BCTN was 4.3- and 4.1-fold higher than that of bare SiO₂/Pt and SiO₂/Pt@PEG-Glu after 12 hours of oral administration, respectively. In addition, although BCTN would predominantly accumulate in the intestines of CRC, BCTN did not cause significant distribution in normal organs compared with other groups (Fig. 6E). However, the targeting effect of BCTN to the CRC-related intestine disappeared when the nanoplatform was preincubated with IgG. Combined with the results of the magnetic beads adsorption experiment in Fig. 3A, it could be concluded that BCTN relied on protein A on SAM to specifically recognize the overexpressed IgG inflammatory intestine of CRC and to efficiently anchor to the intestine of CRC. With superior recognition ability of CRC-related intestine, the SAM camouflage strategy laid a solid foundation for the self-thermophoretic nanoplatform to efficiently penetrate the intestinal barrier of CRC.

Encouraged by the superior intestinal positioning and autonomous penetration performance, we further explored the in vivo multiple barrier penetration ability of the nanoplatform in orthotopic CRC tumor–bearing mice. The in situ intestinal loop model was continued to be used to study the tumor penetration of the nanoplatform. After circulation perfusion in the intestine tract for 5 min, the MFI of SiO₂/Pt and SiO₂/Pt@PEG in the tumor site was only 7.1 ± 2.1 and 19.8 ± 6.4, respectively (Fig. 6, F and G). With the functionalization of PEG-Glu, the MFI of SiO₂/Pt@PEG-Glu in the tumor site was 49.9 ± 12.7, which was significantly higher than that of SiO₂/Pt and SiO₂/Pt@PEG. In contrast, after NIR laser irradiation for 5 min, the MFI of nanoplatform in the tumor site of the SiO₂/Pt@PEG-Glu + NIR group was 147.2 ± 9.0, which was 20.7-, 7.4-, and 3.0-fold higher than that of the SiO₂/Pt, SiO₂/Pt@PEG, and SiO₂/Pt@PEG-Glu groups, respectively. The results showed that massive SiO₂/Pt@PEG-Glu was able to efficiently penetrate the complex matrix after NIR laser irradiation and be rapidly internalized by cancer cells. We unexpectedly found that the fluorescence of SiO₂/Pt@PEG-Glu at the tumor site was evenly dispersed in the tumor site of CRC after NIR laser irradiation. This may be mainly attributed to the self-thermophoresis–facilitated homogeneous distribution of the nanoplatform at the tumor site of CRC, thus also improving the deep penetration at the tumor site of CRC (41).

Furthermore, inductively coupled plasma mass spectrometry (ICP-MS) was applied to examine the distribution of nanoplatform in tumor and major digestive organs involved in oral drug delivery (fig. S28). The ICP-MS results also confirmed the increased distribution of Si content in the tumor site after oral administration. Si contents in major digestive organs were decreased. The above results showed that BCTN could improve the tumor distribution of nanoplatform and avoid the off-target to gastrointestinal tissues.

Last, we stained the tumor cells by anti-Ki67 antibody to evaluate the distribution of the nanoplatform in cancerous cells (fig. S29). The semiquantitative analysis results in fig. S30 showed that BCTN in cancerous cells was significantly higher than that in noncancerous cells, which demonstrated that BCTN had specific recognition ability to cancer cells, thus improving drug delivery efficiency at the cancer cell level.

On the basis of the in vitro and in vivo intestinal barrier penetration experiments, SiO₂/Pt@PEG-Glu could significantly penetrate the intestinal epithelial cells and rapidly reach the CRC cells after NIR laser irradiation. The autonomous movement of SiO₂/Pt@PEG-Glu with effective penetration ability provided a solid foundation for the improvement of drug delivery efficiency in the following in vivo anticancer experiments.

In view of the aforementioned targeting properties of BCTN, blood clearance kinetic experiments were conducted to examine the in vivo circulation properties of nanoplatform. The plasma concentration curves of the nanoplatform were shown in Fig. 6H, and the plasma concentration time profiles were shown in table S1 in the Supplementary Materials. The peak concentration (C_max) of BCTN was 1.9- and 1.5-fold higher than that of SiO₂/Pt and SiO₂/Pt@PEG-Glu, respectively. Besides, the C_max of BCTN after NIR laser irradiation (1.5 W cm⁻²) was 1.40 ± 0.10 μg ml⁻¹, which was 3.3- and 2.6-fold higher than that of SiO₂/Pt and SiO₂/Pt@PEG-Glu, respectively. Furthermore, the area under curve of plasma concentration and half-life of BCTN after NIR laser irradiation in plasma were 3.6- and 1.4-fold higher than that of SiO₂/Pt, respectively. In addition, the bioavailability of BCTN after NIR laser irradiation also increased by 2.6-, 1.8-, and 0.6-fold compared with SiO₂/Pt, SiO₂/Pt@PEG-Glu, and BCTN, respectively, which was consistent with the results in the in vitro and in vivo intestinal mucus barrier penetration experiments.

BCTN vastly improves the in vivo anticancer efficacy

Next, the in vivo anticancer efficacy of BCTN was comprehensively evaluated. The experimental program was carried out according to the schedule in Fig. 7A. CT26-luc cancer cells were injected into mouse cecum to construct an orthotopic colorectal tumor–bearing mouse model on day −13. The tumor-bearing mice received 3 weeks of oral administration with drug-loading nanoplatform twice a week from day 1. The body weight of mice in each group was recorded every 2 days. During the treatment period, the mice in the CP and CP@SiO₂/Pt groups developed significant body weight losses compared with other groups (P < 0.05; Fig. 7B). The symptom of weight loss in mice after chemotherapy was also reported in the literature (42, 43).

Fig. 7. BCTN vastly improves the in vivo anticancer efficacy.

(A) In vivo treatment timeline of CT26-luc orthotopic CRC tumor–bearing mice. (B) Body weight change curve of mice in each group (n = 5). (C) Tumor bioluminescence images of mice in each group. (D) Tumor bioluminescence change curve during the treatment. (E) Tumor mass of mice in each group after 3 weeks of treatment. (F) Tumor metastasis rate of mice after 3 weeks treatment (G) Bioluminescence images of main organs of mice after 3 weeks of treatment (LN, mesenteric draining lymph node). (H) H&E and cleaved caspase-3 staining images of tumor tissues from mice after 3 weeks of treatment. (I) Cleaved caspase-3–positive cell rate of tumor tissues from mice in each group. (J) Changes of flora composition in mice after 3 weeks of treatment (1, normal; 2, saline; 3, CP; 4, CP@SiO₂/Pt; 5, CP@SiO₂/Pt@PEG-Glu; 6, BCTN; 7, BCTN + NIR). (K) β-Analysis of flora composition in mice after 3 weeks of treatment. Values are represented as means ± SD. Significance was calculated via two-way ANOVA (B and D) with Dunnett’s posttest or one-way ANOVA (E, I, and J) with Tukey’s posttest. *P < 0.05, **P < 0.01, and ***P < 0.001.

The chemotherapeutic drug CP always has hepatotoxicity and bone marrow toxicity in clinical settings (44, 45). Routine blood examination showed that the mice in the CP group exhibited significantly decreased white blood cell (WBC), lymphocyte, and monocyte compared with the normal mice (fig. S31). The targeted drug delivery of BCTN could avoid the decrease of WBC in tumor-bearing mice during the treatment period. In addition, BCTN could alleviate the increased ALT and AST caused by CP, indicating that BCTN was able to reduce the hepatotoxicity of CP (fig. S32). In addition, BCTN reduced the bone marrow suppression of CP calculated by the bone marrow cellularity in H&E staining images (figs. S33 and S34).

Bioluminescence imaging was then applied to monitor the tumor growth of tumor-bearing mice in each group. The representative bioluminescence images in Fig. 7C indicated that CP and CP@SiO₂/Pt could not delay the tumor growth of tumor-bearing mice compared with saline. The tumor bioluminescence change curve in Fig. 7D showed that CP@SiO₂/Pt@PEG-Glu exhibited moderate in vivo anticancer effects during the treatment. Encouragingly, BCTN showed a marked inhibiting effect on tumor growth compared with saline (P < 0.01). Furthermore, BCTN with 808-nm laser irradiation (1.5 W cm⁻²) showed the best in vivo anticancer effect in the treatment groups (P < 0.05 compared with BCTN and P < 0.001 compared with CP, CP@SiO₂/Pt, and CP@SiO₂/Pt@PEG-Glu). In addition, the photographs of the tumor dissection harvested in each group after 3 weeks of treatment showed that the mice in BCTN + NIR group had the smallest tumor volume (fig. S35). The tumor in two mice of the BCTN + NIR group disappeared. In addition, BCTN + NIR group produced the best tumor inhibition effect with an inhibition rate of 99.1% after 3 weeks of treatment (Fig. 7E). Moreover, the tumor weight of mice in the BCTN + NIR group after 3 weeks of treatment was significantly lighter than that of the BCTN group (P < 0.05), which confirmed that the NIR laser irradiation was able to improve the treatment efficiency of BCTN on CRC. The above results proved the superior in vivo anticancer efficacy of BCTN after NIR laser irradiation.

The orthotopic colorectal tumor–bearing mice model established in this work could spontaneously form distant metastases. Therefore, the main organs and mesenteric draining lymph nodes of mice in each group were taken for bioluminescence imaging to check the metastases after 3 weeks of treatment (Fig. 7F). The bioluminescence images in Fig. 7G indicated that distant metastases of CRC frequently occurred in the liver and mesenteric draining lymph node in the saline group. Mice in the CP and CP@SiO₂/Pt groups had similar metastasis sites to that of the saline group. Furthermore, pathological images also confirmed the metastases in liver and colorectum in each group (fig. S33).

H&E staining and cleaved caspase-3 (c-caspase-3) immunohistochemistry were used to check the in vivo anticancer effect of nanoplatform (Fig. 7H). The results of H&E staining experiments demonstrated that BCTN and BCTN with NIR laser irradiation induced massive cancer cell death and shrinkage of the nucleus, which proved the remarkable in vivo antitumor effect of BCTN after NIR laser irradiation. The results in c-caspase-3 immunohistochemistry revealed that BCTN with NIR laser irradiation exhibited the highest apoptosis-inducing activity on CRC cells compared to other groups (Fig. 7I).

It is well known that intestinal flora is closely associated with CRC (46, 47). To study the influence of the treatment process on the composition and abundance of intestinal microbiota, 16S ribosomal DNA sequencing technology was applied to analyze the intestinal bacterial composition of mice in each group after 3 weeks of treatment. The results indicated that the mice in the saline and CP groups had not only rapid tumor development but also a significant richness in the pathogenic bacterial abundance such as Campylobacteria and Gammaproteobacteria (Fig. 7J). However, the pathogenic bacterial abundance in the BCTN and BCTN + NIR group reduced after 3 weeks of treatment. In addition, the β-diversity index indicated that the intestinal microbiota of mice in the BCTN + NIR group was closer to normal mice compared with mice in other groups (Fig. 7K), which indicated the superior in vivo anticancer activity of nanoplatform during the treatment process. Together, the nanoplatform not only produced excellent in vivo anticancer capability but also prevented the imbalance of intestinal flora caused by the progression of CRC.

Long-term survival rate after treatment is also an important indicator for cancer therapy efficacy. Therefore, we carried out the in vivo treatment experiment of nanoplatform for 3 weeks (days 1 to 21) and continued to observe for 1 month after treatment to evaluate the long-term survival rate and tumor recurrence possibility (days 22 to 50). The tumor bioluminescence images were shown in fig. S36. After 3 weeks of treatment, the survival rate of mice in the saline group was 80% at day 21. All the mice in the saline group were dead within 50 days. In addition, the mice of CP group were dead within 50 days. In sharp contrast, BCTN + NIR not only significantly delayed the growth of orthotopic CRC tumor but also prolonged the survival rate of mice (fig. S37). The survival rate of mice of BCTN + NIR group at day 50 was 60%. Two of the cured mice in BCTN group had no recognizable tumor metastasis and recurrence during the 50 days of observation period (fig. S38). Therefore, the nanoplatform could significantly improve the survival time and reduce tumor recurrence in orthotopic CRC tumor–bearing mice.

After in vivo anticancer treatment experiments, the degradation of nanoplatform is vital to affect the toxicity to organisms. In this work, the only inorganic component of nanoplatform after treatment was mesoporous silica. Therefore, the contents of Si in different tissues after treatment were detected (fig. S39). The Si content in tumor and colon significantly decreased at 15 days after treatment. At 30 days after treatment, the Si content in main tissues further decreased to less than 10% of the original Si content, which indicated that most nanoplatforms were metabolized and excreted after treatment. The results were consistent with the previous studies on the degradation of silicon-based materials (48–51).

DISCUSSION

Intestinal mucus barrier seriously hindered the delivery efficiency of DDS. Anderski et al. (52) reported polymer poly(lactic-co-glycolic) acid (PLGA) nanoparticles for mucus penetration. However, the penetration of PLGA nanoparticles in biosimilar mucus after incubation of 4 hours was only ~0.38 μm. Müller et al. (53) reported papain-modified nanoparticles to cleave mucoglycoprotein substructures of mucus. Although this strategy contributed to the diffusion of DDS, the intestinal epithelial cells were directly exposed to the complex intestinal microenvironment containing numerous bacteria and viruses. Therefore, there is still lack of efficient and nondestructive method to cross the intestinal mucus barrier.

Nanomotors have attracted tremendous attention in recent years because of their promising applications in the field of complex matrix penetration (54), but classical nanomotors lack biological chemotaxis to colonize the complex physiological barrier, such as CRC-related intestine. Wan et al. (55) camouflaged nanomotors with platelet membrane to realize the effective inflammatory chemotactic aggregation in thrombus sites. In this work, BCTN after bacterial biomimetic membrane coating had specific recognition ability to the intestinal inflammatory environment of CRC and with no significant effect on the highly efficient movement performance of BCTN. In vivo imaging experiment demonstrated that the aggregation ability of nanoplatform on CRC-related intestines was 4.3-fold higher than that of the uncoated nanoplatform after 12 hours of oral administration. Therefore, the SAM camouflage strategy of bacterial biomimetic membrane laid a solid foundation for self-thermophoretic technology to penetrate the multiple intestinal barrier.

The intestinal mucus barrier penetration experiments demonstrated that the penetration depth of the nanoplatform was 67.9 ± 8.7 μm after 5-min laser irradiation, which was much higher than that of passive diffusion nanoparticles (15.2 ± 4.4 μm). Different from previous studies, this strategy of autonomous movement across the intestinal mucus barrier not only produced highly efficient penetration effect but also nondestructively penetrated the dense and viscous intestinal mucus barrier. The improved bioavailability is beneficial not only for effective treatment of orthotopic colorectal tumor–bearing mice but also for inhibiting the growth of CRC metastases in the following in vivo anticancer experiments (56).

In vivo anticancer efficacy experiments showed that the nanoplatform effectively inhibited the tumor growth in CRC tumor–bearing mice, alleviated the side effects of chemotherapeutic drugs, and reduced the cancer metastases. Metastasis is one of the leading causes of death in patients with CRC. Approximately 50% of patients with CRC have liver metastases (57, 58). Notably, BCTN and BCTN + NIR laser irradiation not only exhibited a significant reduction in orthotopic tumor volume but also inhibited tumor metastasis rate of tumor-bearing mice. Enhanced drug delivery was able to lead to superior chemotherapeutic effects. The superior chemotherapy effect may also cause a significant decrease in tumor metastasis through immune activation and immunogenic cell death in the tumor site (59). The 3 weeks of oral administration of BCTN also alleviated the intestinal flora imbalance induced by CRC, which indicated the superior in vivo anticancer activity of the nanoplatform during the treatment process.

In summary, we developed a BCTN-augmented CRC therapy through precise intestinal positioning and autonomous mucus penetration. SAM camouflage endowed the nanoplatform with biological chemotactic properties toward the intestinal site of CRC. The nanoplatform exhibits excellent penetration performance on the in vitro and in vivo intestinal mucus barrier after the accelerated accumulation at the intestinal site of CRC. Owing to the precise intestinal positioning and self-thermophoresis–inspired autonomous mucus penetration, the in vivo intestinal mucus barrier penetration efficiency of BCTN after NIR laser irradiation was vastly improved. Moreover, BCTN with NIR laser irradiation exhibited a tumor inhibition rate of 99.1% on orthotopic CRC tumor–bearing mice, which provides new insight into nondestructive penetration of complex biological barriers and effective oral targeted therapy of CRC.

MATERIALS AND METHODS

Study design

The objective of this study was to develop BCTN-augmented CRC therapy through autonomous mucus penetration. First, we prepared BCTN and characterized the biological chemotaxis and autonomous movement characteristics of the nanoplatform. Second, mucus-producing transwell model and mouse colon were used to evaluate the intestinal barrier penetration performance of the nanoplatform in vitro and in vivo. Then, BCTN investigated the in vitro anticancer activity to Caco-2 cell monolayer model and mucus-producing transwell model. After that, in vivo distribution experiments were conducted to examine the targeting performance of nanoplatform to CRC intestinal segments. Last, we assessed the in vivo anticancer performance of nanoplatform by the tumor bioluminescence change during treatment and tumor weight of mice after treatment.

Preparation of mesoporous silica/platinum (SiO₂/Pt)

The MSN with a diameter of ~50 nm was prepared according to previous method with slight modifications (60). Briefly, hexadecyltrimethylammonium bromide (50 mg) and NaOH (6 mg) were dissolved in deionized water (20 ml) and heated to 80°C. Then, tetraethyl orthosilicate (0.25 ml) was added drop by drop into the mixture. After reaction at 80°C for 2 hours, the mixture was washed with water and 1% NaCl/MeOH solution several times to remove the template. The as-prepared MSN was dispersed in water, dropped on silicon wafer to form a monolayer structure, and dried naturally. Subsequently, platinum was deposited on the surface of MSN using a vacuum evaporator to obtain SiO₂/Pt. The SiO₂/Pt were scraped off the silicon wafer gently and dispersed evenly by ultrasound.

PEG functionalization of SiO₂/Pt

SiO₂/Pt (20 mg) were dispersed in N,N′-dimethylformamide (DMF; 2 ml), and 3-aminopropyltrimethoxysilane (2 μl) was added to functionalize nanoparticles with amino group. Subsequently, amino-functionalized SiO₂/Pt (20 mg) were dispersed in DMF (2 ml); then, 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (EDC; 2 mg), N-hydroxysuccinimide (NHS; 2 mg), and PEG₂₀₀₀-SS-COOH (2 mg) were added. The reaction mixture was reacted under magnetic stirring overnight and then washed with DMF and water several times to obtain PEG-functionalized SiO₂/Pt (SiO₂/Pt@PEG).

Glucose functionalization of SiO₂/Pt@PEG

SiO₂/Pt@PEG (20 mg) were dispersed in DMF (2 ml), and EDC (2 mg), N,N-dimethylformamide (0.2 mg), and succinic anhydride (2 mg) were added to functionalize with carboxyl group. The reaction mixture was washed with DMF and water several times. Subsequently, SiO₂/Pt@PEG (20 mg) was dispersed in DMF (2 ml); then, EDC (2 mg), NHS (2 mg), and glucose (1 mg) were added. The reaction mixture was reacted under magnetic stirring overnight and then washed with DMF and water several times to obtain glucose-functionalized SiO₂/Pt@PEG (SiO₂/Pt@PEG-Glu).

Drug loading of SiO₂/Pt@PEG-Glu

CP (5 mg) was dissolved in DMF (1 ml), and SiO₂/Pt@PEG-Glu (5 mg) was added. After stirring overnight, the mixture was centrifuged at 12,000 rpm for 10 min. The precipitate was CP-loading SiO₂/Pt@PEG-Glu (CP@SiO₂/Pt@PEG-Glu). The content of free CP in supernatant was determined by high-performance liquid chromatography. The loading content was calculated as [(original CP content − free CP content)/(original nanoparticle content + original CP content − free CP content)] × 100% (i). The encapsulation efficiency was calculated as [(original CP content − free CP content)/original CP content] × 100% (ii).

Bacterial biomimetic membrane camouflage of CP@SiO₂/Pt@PEG-Glu

The procedure was according to our previous method (24). Briefly, S. aureus (American Type Culture Collection, 29213) was cultured in Mueller-Hinton broth. S. aureus culture medium (1 ml) was harvested and washed with PBS (pH 7.4). Then, lysozyme [1 mg ml⁻¹ in tris buffer (pH 8.0), 100 μl] was added and shaken at 37°C for 20 min. The reaction mixture was centrifuged at 5000 rpm with an ultrafiltration centrifuge tube (100 kDa; Corning) and washed with PBS (pH 7.4) to obtain bacterial SAM. Subsequently, CP@SiO₂/Pt@PEG-Glu was mixed with SAM and extruded through 400-, 200-, and 100-nm polycarbonate porous membranes for five cycles with an Avanti mini extruder (Avanti Polar Lipids). The mixture was centrifuged using an ultrafiltration centrifuge tube at 5000 rpm for 10 min and concentrated by nitrogen blowing to obtain BCTN.

Magnetic beads adsorption experiments

SiO₂/Pt@PEG-Glu and BCTN were labeled with fluorescein isothiocyanate (FITC), and then, FITC-labeled BCTN was incubated with IgG to prepare FITC-loaded BCTN + IgG. The magnetic beads (Yeasen Biotechnology Co. Ltd., Shanghai, China) were functionalized with IgG and then mixed with FITC-loaded nanoparticles (1 mg ml⁻¹, 200 μl) for 30 min. The magnetic beads were precipitated by magnets and washed three times with PBS (pH 7.4). The fluorescence intensity of magnetic beads was detected by fluorescence spectrophotometer.

Intestinal penetration experiments

SiO₂@PEG-Glu and SiO₂/Pt@PEG-Glu were loaded with Cy5.5. Sprague Dawley rats (220 to 280 g) were sacrificed to take the colorectum. Then, Cy5.5-labeled SiO₂@PEG-Glu and SiO₂/Pt@PEG-Glu were incubated with colorectum for 5 min with or without 808-nm laser (1.5 W cm⁻²) irradiation. Then, intestinal mucus was stained by FITC-labeled wheat germ agglutinin. After the colorectum was fixed on the glass slide, the fluorescence was detected by CLSM. The penetration depth of nanoparticles was calculated using LAS AF Lite 3.3.0.

Transepithelial transport experiments

Mixed Caco-2 and HT29-MTX (7:3) cells were seeded in a transwell apical chamber fitted with polycarbonate membranes (12-well, 3.0 μm; Corning). When the transepithelial electrical resistance was higher than 300 ohms cm⁻², CT26 cells were seeded in the basal chamber and incubated for 12 hours. SiO₂/Pt, SiO₂/Pt@PEG, and SiO₂/Pt@PEG-Glu were loaded with FITC. FITC-loaded SiO₂/Pt, SiO₂/Pt@PEG, and SiO₂/Pt@PEG-Glu (50 μg ml⁻¹, in terms of SiO₂/Pt) were added into the apical chamber of a transwell plate for 3 hours. The transwell plate in SiO₂/Pt@PEG-Glu + NIR group received 808-nm laser irradiation (1.5 W cm⁻²) for 5 min after the addition of nanoparticles. In addition, dapagliflozin (12.5 μM) was added in the medium of the apical chamber in the SiO₂/Pt@PEG-Glu + dapagliflozin group to the inhibit glucose transporter of Caco-2 and HT29-MTX cells. Then, the cells were washed with PBS (pH 7.4) and detected by CLSM. The internalization of cells was further analyzed by flow cytometry (BD LSRFortessa; BD, San Jose, USA).

Transepithelial transport anticancer experiments

Mixed Caco-2 and HT29-MTX (7:3) cells were seeded in a transwell apical chamber fitted with polycarbonate membranes (12-well, 3.0 μm; Corning). When the transepithelial electrical resistance was higher than 300 ohms cm⁻², CT26 cells was seeded in the basal chamber and incubated for 12 hours. CP, CP@SiO₂/Pt, and CP@SiO₂/Pt@PEG-Glu (5 μg ml⁻¹, in terms of CP) were added into the apical chamber and incubated for 3 hours. The cells in the CP@SiO₂/Pt@PEG-Glu + NIR group received 808-nm laser irradiation (1.5 W cm⁻²) for 5 min after the addition of nanoparticles. Then, CT26 cells were washed with PBS (pH 7.4) and stained by an annexin V–FITC apoptosis detection kit (Beyotime Biotechnology, Shanghai, China). The cells were further analyzed by flow cytometry (BD LSRFortessa; BD, San Jose, CA).

Establishment of orthotopic CRC animal model

The establishment of the orthotopic CRC animal model was according to previous work with slight modifications (37). The mice were anesthetized by isoflurane. After being disinfected with an alcohol swab on the abdomen, the mice received a median incision on the lower abdomen to exteriorize the cecum. CT26-luc cells were diluted to 4 × 10⁶ cell ml⁻¹ by mixing Matrigel (10 μg μl⁻¹; Corning) and RAPI 1640 medium (3:1, v/v). Then, the cells were injected into the cecal wall by a 30-gauge needle (Hamilton Company, Bonaduz, Switzerland). Then, the cecum was returned to the abdominal cavity, and the abdominal cavity was sutured with 5-0 absorbable suture. All animal experiments were approved by the Life Sciences Ethical Review Committee of Zhengzhou University (Zhengzhou, China).

In vivo intestinal barrier penetration experiments

The experiment procedure was performed as previously described with slight modifications (38). SiO₂/Pt, SiO₂/Pt@PEG, and SiO₂/Pt@PEG-Glu were labeled with NIR fluorophore Cy5.5. Fourteen days after orthotopic tumor cell inoculation, the mice were fasted for 12 hours with free access to water and then were anesthetized with sodium pentobarbital. The intestinal loop was exposed, and latex tubing was inserted into the upper portion of the intestine, and then, it was connected to a peristaltic pump. The peristaltic pump was used to perfuse the circulating solution, which contained the Cy5.5-labeled SiO₂/Pt, SiO₂/Pt@PEG, and SiO₂/Pt@PEG-Glu. The other end of the latex tubing was then inserted into the lower portion of the intestine to form a circuit. The mice in SiO₂/Pt@PEG-Glu + NIR group received 808-nm laser irradiation (1.5 W cm⁻²) for 5 min. Five minutes later, the intestines of tumor site were taken and measured by TEM and CLSM.

In vivo antitumor experiments

Fourteen days after orthotopic tumor cell inoculation, the mice received intraperitoneal injection of d-luciferin (150 mg kg⁻¹; J&K Chemical Ltd., China) to check the bioluminescence intensity of the tumor. The mice with a bioluminescence intensity of ~1 × 10⁶ photons (p) s⁻¹ cm⁻² sr⁻¹ were used for in vivo anticancer experiments. Then, the mice were randomly divided into six groups (n = 5) and received intragastric administration of saline (control), CP, CP@SiO₂/Pt, CP@SiO₂/Pt@PEG-Glu, and BCTN (3 mg kg⁻¹, in terms of CP) twice a week for 3 weeks. Twelve hours after intragastric administration, the mice in BCTN + NIR group received 808-nm laser irradiation (1.5 W cm⁻²) for 5 min. The body weight of mice in each group was recorded every 2 days. Furthermore, the mice in each group received intraperitoneal injection of d-luciferin (150 mg kg⁻¹; J&K Chemical Ltd., China) every 4 days; then, the mice were anesthetized and imaged under the small animal imaging system (Caliper Life Sciences Inc., USA) to monitor the tumor bioluminescence. After 3 weeks of treatment, the mice in each group were sacrificed to take the blood, tumor, and main tissues. Then, the blood was centrifuged at 3000 rpm for 5 min for serum chemical analysis. In addition, mesenteric draining lymph nodes were harvested for bioluminescent imaging to check the metastasis of tumor cells. The H&E staining images of femur were analyzed for total bone marrow cellularity by ImageJ software. Furthermore, H&E staining and c-caspase-3 immunohistochemistry assay were performed on the tumors and lungs to examine the antitumor effect of the nanoplatform.

Supplementary Materials

This PDF file includes:

Supplementary Text

Figs. S1 to S39

Table S1

References

Other Supplementary Material for this manuscript includes the following:

Movies S1 to S3

View/request a protocol for this paper from Bio-protocol.