pH-responsive selenium nanoplatform for targeted drug release and immune remodeling in microbiota-associated colorectal cancer

Abstract

Fusobacterium nucleatum (F. nucleatum) is implicated in colorectal cancer (CRC) initiation, progression, and resistance to therapy, while conventional antibiotics show poor specificity and disrupt intestinal homeostasis. Here, we present a dual bacteria–tumor elimination strategy using a multifunctional nanoagent, SeNPs@CBT, which integrates antibacterial, antitumor, and tumor-targeting functions. Selenium nanoparticles serve as carriers, coated with chitosan-4-carboxyphenylboronic acid and loaded with caffeic acid phenethyl ester (CAPE) via pH-sensitive borate ester bonds. In the acidic tumor microenvironment, SeNPs@CBT release CAPE to eradicate F. nucleatum and modulate immune responses, while the exposed phenylboronic acid groups enhance tumor recognition and uptake, synergistically inducing apoptosis. In an F. nucleatum–associated CRC mouse model, SeNPs@CBT markedly suppressed tumor growth and exhibited potent synergistic effects in bacterial clearance, tumor cell killing, and immune activation. Mechanistic and transcriptomic analyses revealed induction of cell cycle arrest, mitochondrial dysfunction, and activation of the p53 signaling pathway. Together, SeNPs@CBT, based on the triple synergy of bactericidal, tumoricidal, and immune activation, represents a promising nanomedicine platform for the precise treatment of microbiota-related CRC, particularly F. nucleatum–enriched subtypes.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-026-04229-0.

Introduction

With a deeper understanding of the mechanisms driving cancer initiation and progression, it has become increasingly clear that cancer is not solely a consequence of uncontrolled proliferation of malignant cells, but rather a complex and dynamic disease influenced by the tumor microenvironment (TME) [1]. TME comprises a dynamic and heterogeneous ecosystem composed of stromal cells, immune components, and, importantly, microorganisms, which collectively regulate tumor initiation, progression, invasion, and metastasis [2]. In recent years, growing evidence has highlighted the significant role of tumor-associated microbiota in modulating the TME and influencing cancer progression and treatment outcomes [3]. Among these, F. nucleatum, an opportunistic pathogen that enriched in the gut mucosa of CRC patients, has been strongly associated with the development of CRC [4]. Its abundance is markedly elevated in CRC tissues compared to healthy controls, and it plays a role in tumor progression by activating β-catenin signaling, upregulating oncogene expression, and creating a pro-inflammatory environment [5]. Additionally, F. nucleatum impairs anti-tumor immunity by inhibiting T cells and natural killer (NK) cells, promoting immune evasion, enhancing resistance to chemotherapy, and accelerating tumor metastasis. Given the multifaceted pro-tumorigenic effects associated with these pathogens, the targeted elimination of tumor-associated pathogenic bacteria, particularly F. nucleatum, has emerged as a promising adjunctive approach in the treatment of CRC [6–8].

Although antibiotic therapy provides a direct approach to modulate tumor-associated microbiota [9], prolonged or broad-spectrum administration may contribute to the development of antimicrobial resistance and disturb the balance of the commensal microbiome, which could result in systemic adverse effects [10–12]. Nanotechnology-based drug delivery systems enable precise therapeutic targeting and significantly reduce off-target effects [13]. A variety of nanocarriers, including mesoporous silica nanoparticles [14], liposomes [15], and polymeric micelles [16], have been investigated for their application in cancer therapy, utilizing the enhanced permeability and retention (EPR) effect to achieve effective tumor accumulation. However, these systems often exhibit limited intrinsic therapeutic efficacy and delayed biodegradation. For instance, gold nanoparticles may remain in the liver for several months [17], whereas organically coated nanodots can persist in the body for years [18], which raises concerns regarding long-term biosafety.

To overcome these challenges, we developed a selenium nanoparticle (SeNP)-based multifunctional nanodrug delivery platform that integrates antibacterial, antitumor, and tumor-targeting capabilities. Selenium, an essential trace element, possesses antioxidant, immune-modulating, and anticancer properties [19, 20]. Notably, SeNPs demonstrate higher bioavailability, reduced toxicity, and enhanced biocompatibility compared to other selenium formulations [21]. More importantly, SeNPs possess intrinsic anticancer activity, allowing them to act not only as drug carriers but also as active therapeutic agents, thereby addressing a key limitation of conventional inert nanocarriers such as silica nanoparticles, liposomes, and polymeric micelles [22–24]. On this basis, the tunable size and redox-related properties of SeNPs support their application as a versatile drug delivery nanoplatform, enabling efficient loading and controlled delivery of various chemotherapeutic agents or bioactive molecules to improve therapeutic outcomes [25, 26]. However, unmodified SeNPs generally exhibit limited tumor specificity, highlighting the necessity of surface functionalization for targeted delivery. Accordingly, diverse functional ligands have been introduced onto SeNP surfaces to achieve selective recognition of tumor-associated targets, including small molecules (e.g., folic acid, lactic acid, and galactose) that promote receptor-mediated endocytosis, polysaccharides and polymers such as chitosan and hyaluronic acid that enhance biocompatibility and target receptors like CD44, as well as peptides and antibodies that further improve targeting specificity [27–37]. Collectively, these functionalization strategies demonstrate the versatility of SeNPs as targeted nanodrug delivery systems and highlight the necessity of rational ligand selection for the construction of multifunctional SeNP-based platforms in cancer therapy.

In this work, we designed a novel nanoplatform, CAPE-CPBA-CTS@SeNPs (hereafter referred to as SeNPs@CBT), in which chitosan (CTS) functionalized with 4-carboxyphenylboronic acid (CPBA) acts not only as a morphological stabilizer but also as a functional ligand (Fig. 1). The CPBA moiety provides dual functionalities: (1) selective binding to cis-diol-containing molecules, which enable pH-responsive drug loading and release [38]; and (2) specific recognition of sialic acid residues that are overexpressed on the surface of CRC cells, thereby enhancing tumor-targeting capability [39]. The natural antibacterial agent caffeic acid phenethyl ester (CAPE) was covalently conjugated to the system via reversible borate ester bonds with CPBA, ensuring pH-sensitive release under the mildly acidic conditions typical of the TME (pH 6.5 - 6.9) [40]. This dual responsiveness ensures that CAPE is selectively released within tumor sites, maximizing antibacterial and antitumor effects while sparing healthy tissues. Compared to traditional nanocarriers, SeNPs@CBT presents several distinct advantages. First, the intrinsic anticancer activity of SeNPs complements the antibacterial action of CAPE, enabling a synergistic inhibition of both tumor cells and pathogenic bacteria, such as F. nucleatum. Second, phenylborate groups confer active targeting through the recognition of sialic acid, extending beyond the passive EPR effect utilized by conventional carriers. Third, unlike inorganic nanoparticles, which are prone to bioaccumulation, SeNPs exhibit efficient metabolic clearance and exhibit minimal long-term toxicity.

Fig. 1.

Schematic illustration of the synthetic strategy and tumor inhibition mechanism of SeNPs@CBT. CPBA capable of targeting cancer cells, was covalently grafted onto CTS via amide bond formation to yield a functionalized capping agent. SeNPs@BT were subsequently synthesized using CPBA-modified CTS as both stabilizer and targeting moiety. The natural antibacterial compound CAPE was then conjugated onto the nanoparticle surface via reversible borate ester bonds, forming the final nanoplatform SeNPs@CBT. Upon intravenous administration, SeNPs@CBT preferentially accumulate at tumor sites through both the EPR effect and active targeting mediated by phenylborate-sialic acid interactions. In the acidic TME, the borate ester bonds are cleaved, triggering the controlled release of CAPE, which selectively eradicates F. nucleatum residing within the tumor niche. Bacterial clearance reshapes the immunosuppressive microenvironment and restores anti-tumor immune surveillance. Concurrently, the SeNPs@BT core exerts intrinsic cytotoxic effects against tumor cells. This combinatorial approach, integrating targeted bacterial elimination, direct tumor cell killing, and immune activation, achieves synergistic suppression of tumor growth and metastasis, offering a promising strategy for effective CRC therapy

Our experimental data demonstrate that SeNPs@CBT effectively eradicates F. nucleatum, suppresses CRC progression, reactivates antitumor immune responses, and achieves enhanced therapeutic efficacy without causing damage to major organs. As a multifunctional and biodegradable nanoplatform, it presents a promising approach for targeting the interconnected microbial and oncogenic mechanisms underlying CRC, thereby contributing to the development of next-generation nanotherapeutics.

Materials and methods

Reagents and chemicals

3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) and CPBA were purchased from Aladdin (Shanghai, China). The BacLight™ Bacterial Membrane Potential Kit was obtained from Thermo Fisher Scientific (Massachusetts, USA). H₂DCFDA (DCFH-DA) was purchased from Sangon Biotech (Shanghai, China). The Bacterial Viability Assay Kit and Cell Cycle Detection Kit were purchased from Beyotime Biotechnology (Shanghai, China). L(+)-Ascorbic acid was obtained from Kermel (Tianjin, China), sodium selenite from Alfa Aesar (Ward Hill, MA, USA), chitosan (molecular weight ≈ 30,000 Da) from Macklin (Shanghai, China), and N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride (EDC) from Aladdin (Shanghai, China).

Cell lines and culture conditions

The human cancer cell lines HCT116(CVCL_D4CW), HT-29(CVCL_0320), KYSE-150(CVCL_1348), SW579(CVCL_3603), AGS(CVCL_0139), and A375(CVCL_0132), as well as the immortalized normal cell lines HL7702(CVCL_6926) and HcoEpiC(CVCL_2379), were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA). Cells were cultured according to ATCC recommendations in Dulbecco’s Modified Eagle Medium (DMEM, basic formulation; Gibco, Grand Island, NY, USA) or Roswell Park Memorial Institute (RPMI) 1640 medium (Gibco), supplemented with 10% fetal bovine serum (FBS; Gibco) and 1% penicillin–streptomycin (Gibco). All cells were maintained at 37 °C in a humidified incubator with 5% CO₂. Cell lines were tested for mycoplasma contamination every 4–6 weeks using a PCR-based detection kit (MycoAlert™, Lonza) following the manufacturer’s instructions. Only mycoplasma-negative cells were used in experiments.

Preparation of SeNPs@CBT

Dissolve 321.30 mg of CTS in 50 mL of 0.3% (v/v) acetic acid solution and stir with a magnetic stirrer. In parallel, 414.85 mg of CPBA and 250 mg of N-hydroxysuccinimide (NHS) were dissolved in methanol and stirred at room temperature for 30 min. Then, 333 mg of 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC) and the CTS solution were added to the CPBA/NHS solution, and the reaction mixture was stirred for 24 h at room temperature. The resulting solution was dialyzed using a dialysis membrane with a molecular weight cut-off (MWCO) of 1,000 Da for 72 h to remove unreacted small molecules. The final product, termed BT, was collected by freeze-drying and stored at 4 °C for further use.

To synthesize selenium nanoparticles, 50 mL of 50 mM sodium selenite (Na₂SeO₃) solution was added to a round-bottom flask, followed by the addition of 70 mg of BT. The mixture was stirred for 30 min at room temperature to allow for BT-mediated surface modification. Next, 20 mL of 500 mM L-ascorbic acid was slowly introduced as the reducing agent, and the reaction was allowed to proceed for 4 h under continuous stirring. The resulting SeNPs@BT colloidal solution was dialyzed against deionized water using a 10,000 Da MWCO membrane for 72 h to remove residual reactants and byproducts.

Dissolve 20 mg of CAPE in 100 μL of dimethyl sulfoxide (DMSO), and then slowly added to 20 mL of the SeNPs@BT colloidal solution under gentle stirring. The pH of the mixture was adjusted to 7.5 to facilitate the formation of borate ester bonds between CAPE and phenylboronic acid groups on BT. The mixture was stirred at room temperature for 4 h. After completion, the suspension was centrifuged at 13,000 rpm for 20 min, and the pellet was resuspended in 20 mL of PBS (pH 7.5) to yield the final SeNPs@CBT colloidal formulation with a mass concentration of 2.79 mg/mL.

Characterization of SeNPs@CBT

The hydrodynamic size distribution and zeta potential of the nanoparticles were determined using a nanoparticle analyzer (Zetasizer Pro, Malvern Instruments, U.K.). Surface morphology was assessed by scanning electron microscopy (SEM, HITACHI S-4800) and transmission electron microscopy (TEM, JEOL JEM-2100). The elemental composition and surface chemical states were analyzed using X-ray photoelectron spectroscopy (XPS, Thermo Scientific K-Alpha). Fourier-transform infrared (FTIR) spectra were collected using a BFRL Rayleigh WQF-530 spectrometer to confirm chemical functionalities. The molecular structure of the synthesized BT conjugate was characterized by proton nuclear magnetic resonance (¹H NMR) spectroscopy on a Bruker Avance III™ HD spectrometer (400 MHz).

Quantitative analysis of free CPBA on SeNPs@CBT surface

The amount of free CPBA on the surface of SeNPs@CBT was quantified using an Alizarin Red S (ARS) fluorescence probe assay. Briefly, CPBA standard solutions at concentrations of 0, 0.05, 0.1, 0.25, and 0.5 mM were prepared in PBS buffer (pH 7.5). Each CPBA solution was mixed with a 1 mM ARS solution at a volume ratio of 1:1 and allowed to react at room temperature for 10 min in the dark. Fluorescence spectra were recorded with an excitation wavelength of 460 nm and an emission range of 500–750 nm to construct the calibration curve correlating fluorescence intensity with CPBA concentration.

SeNPs@CBT samples were prepared at a concentration of 1 mg/mL and incubated with ARS under the same conditions prior to fluorescence measurement. To correct for potential interference arising from nanoparticle-induced light scattering and non-specific adsorption of ARS, blank nanoparticles without CPBA grafting were treated identically and used for background subtraction.

pH- responsive release of CAPE

To construct the standard calibration curve, CAPE was dissolved in PBS to prepare solutions at concentrations of 0, 1, 3, 5, 7, and 10 μg/mL, and the absorbance was measured using a UV–Vis spectrophotometer (λ=325 nm). For drug release evaluation, 1 mL of SeNPs@CBT solution (1 mg/mL) was sealed in a dialysis bag (MWCO = 10 kDa) and immersed in 40 mL of phosphate-buffered saline (PBS) at pH 4.5 or 7.4. The samples were maintained at 37 °C under constant shaking. At predefined time points (0, 0.5, 1, 2, 4, 6, 8, 10, 12, 24, 36, 48, 60, and 72 h), 2 mL of the release medium was collected and replaced with an equal volume of fresh PBS to maintain sink conditions. The CAPE content in each sample was quantified via UV–Vis spectroscopy using the calibration curve. The cumulative release percentage of CAPE was calculated using the following formula:

here Ve represents the sampling volume (2 mL), V₀ denotes the initial volume (40 mL), C_i and C_n refer to the concentrations of CAPE at the ith and nth sampling time points, respectively, and m_CAPE indicates the total amount of CAPE encapsulated in SeNPs@CBT.

Assessment of antibacterial efficacy and mechanism in vitro

F. nucleatum (ATCC 23726), was cultured anaerobically in tryptic soy broth (TSB) medium supplemented with 1% Bacto™ Peptone, 3% modified TSB (mTSB), and 20 mM L-cysteine. The culture was purged with CO₂ and incubated at 37 °C in an anaerobic chamber until the optical density at 600 nm (OD₆₀₀) reached 0.6. CAPE, SeNPs@BT, and SeNPs@CBT were diluted to the designated concentrations using fresh TSB medium. The bacterial suspension was adjusted to 8 × 10⁸ CFU/mL. For each treatment, 50 μL of bacterial suspension and 50 μL of drug solution were added to 96-well plates. Plates were incubated anaerobically at 37 °C for 8 h. After treatment, 50 μL aliquots were serially diluted, plated on TSPC agar (composed of 1% Bacto™ Peptone, 3% mTSB, 1.5% Bacto™ Agar, and 20 mM L-cysteine), and incubated anaerobically for 48 h. The number of colony-forming units (CFUs) was counted to assess bacterial viability.

Following treatment, bacterial morphology was observed by scanning electron microscopy (SEM) after fixation with 2.5% glutaraldehyde and graded ethanol dehydration. Bacterial viability was assessed by live/dead fluorescent staining using a commercial kit. Intracellular reactive oxygen species (ROS) levels and membrane potential changes were measured following the respective manufacturer’s protocols.

Cell culture

Human colorectal cancer cell lines (HCT116 and HT-29) were cultured in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin–streptomycin. Normal human hepatocytes (HL7702) and human colonic epithelial cells (HCoEpiC) were maintained in RPMI-1640 medium containing 10% FBS and 1% penicillin–streptomycin. All cell lines were incubated at 37 °C in a humidified atmosphere containing 5% CO₂.

Quantification of intracellular and extracellular F. nucleatum

To assess the antibacterial efficacy of different treatments against both intracellular and extracellular F. nucleatum, HCT116 cells were seeded at a density of 1 × 10⁵ cells per well in 48-well plates and allowed to adhere for 12 h. The culture medium was then removed and replaced with 500 μL of antibiotic-free DMEM containing F. nucleatum at 2 × 10⁷ CFU/mL. After 2 h of co-incubation to facilitate bacterial attachment and internalization, drug treatments were applied for an additional 8 h. Subsequently, the supernatant was collected into centrifuge tubes, and each well was washed once with 500 μL of PBS. The wash was combined with the previously collected supernatant and designated as the extracellular bacterial suspension. For quantification of intracellular bacteria, 500 μL of ultrapure water was added to each well and incubated for 10 min to lyse the host cells. The resulting lysate was collected, and each well was rinsed once with 500 μL of PBS; this rinse was pooled with the lysate and referred to as the intracellular bacterial suspension. Both types of bacterial suspensions were serially diluted and plated on TSPC agar plates. Colony-forming units (CFU) were counted after 48 h of anaerobic incubation at 37 °C.

Cytotoxicity assay

The cytotoxic effects of CAPE, SeNPs@BT, and SeNPs@CBT were evaluated in both cancerous and normal cell lines using the MTT assay. HCT116, HT-29, HL7702, and HCoEpiC cells were seeded in 96-well plates at a density of 1 × 10⁴ cells per well and incubated for 12 hours. The culture medium was then replaced with fresh medium containing the respective test compounds, followed by an additional 24-h incubation period. After treatment, the medium was removed and replaced with 100 μL of MTT solution (100 μg/mL). After 2 h of incubation at 37 °C, the resulting formazan crystals were solubilized with 100 μL of DMSO. The absorbance at 490 nm was measured using a microplate reader. Cell viability was calculated using the following equation:

Cells treated with drug-free medium served as the negative control (100% viability). All experiments were performed in triplicate, and results are expressed as the mean ± standard deviation (SD).

Cellular uptake, intracellular release, and self-monitoring localization in vitro

To evaluate cellular uptake and intracellular localization of SeNPs@CBT, eight cell lines including HCT116, HT-29, KYSE-150, SW579, AGS, A375, HL7702, and HCoEpiC were seeded in 6-well plates at a density of 1 × 10⁶ cells per well and cultured for 12 hours. DiI-labeled SeNPs@CBT were added to the culture medium and incubated for an additional 6 hours. Cells were then harvested via trypsinization, washed with PBS, and collected into centrifuge tubes for flow cytometry analysis using a BD Accuri™ C6 Plus (BD Biosciences, USA).

To visualize intracellular distribution, HCT116 cells in the logarithmic growth phase were seeded into confocal imaging dishes at a density of 5 × 10⁵ cells per well and incubated for 12 hours. DiI-labeled SeNPs@CBT were introduced at a final concentration of 3 μg/mL. At designated time points (3, 6, 12, and 24 hours), cells were washed with PBS, fixed with 4% paraformaldehyde, and stained with DAPI for nuclei and FITC-conjugated membrane dye. Fluorescent images were acquired using a laser scanning confocal microscope (Olympus FV3000, Japan) to assess uptake, release, and subcellular localization of the nanoparticles.

Animal studies

Female BALB/c nude mice (3–4 weeks old) and KM mice (5–6 weeks old) were obtained from Liaoning Changsheng Biotechnology Co., Ltd. (Dalian, China). Mice were housed in a pathogen-free environment with controlled temperature and humidity, and all procedures followed the regulations of the China Animal Protection Law (CAPL).

In vivo antitumor efficacy evaluation

HCT116 tumor-bearing mice were established via subcutaneous injection of a 1:1 mixture of HCT116 cell suspension and Corning® Matrigel® Matrix into the axillary region (100 μL per mouse). When tumor volumes reached approximately 180–200 mm³, animals were randomly divided into five groups (n = 6 per group). In groups designed to model F. nucleatum-associated tumors, mice received intratumoral injections of F. nucleatum (2 × 10⁷ CFU/mL, 100 μL total volume) at multiple sites one day prior to drug administration. Therapeutic agents (PBS, SeNPs@BT, or SeNPs@CBT) were administered via tail vein injection every other day for a total of six doses (5 mg/kg body weight per dose). Tumor volumes and body weights were monitored bi-daily. Tumor volume was calculated using the formula: Tumor volume = a × b²/2, where a is the tumor length and b is the width. At the endpoint, mice were sacrificed and tumors were excised, weighed, and fixed in 4% paraformaldehyde for subsequent histological and molecular analysis.

For clarity, we state that the direct, multi-site intratumoral injection of F. nucleatum into subcutaneous tumors was employed to establish a controlled local model to evaluate interactions among the microorganism, the tumor, and the nanoparticle platform. Although this approach does not fully recapitulate the native tumor microenvironment, it minimizes inter-animal variability and allows precise control over the timing and dose of exposure to a defined bacterial species.

Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay for tumor apoptosis

Tumor tissues were fixed in 4% paraformaldehyde for at least 24 hours, followed by gradient dehydration in ethanol, clearing with xylene, and paraffin embedding. Sections were deparaffinized, rehydrated, and re-fixed in paraformaldehyde before treatment with proteinase K and permeabilization buffer. A labeling mixture consisting of terminal deoxynucleotidyl transferase (TdT), dUTP, and reaction buffer was freshly prepared at a ratio of 2:5:50. Tissue sections were covered with the labeling solution and incubated at 37 °C for 1 hour. Nuclei were counterstained with DAPI. Apoptotic cells were visualized using a fluorescence microscope (Nikon Eclipse C1, Japan), and images were acquired for quantitative analysis.

Enzyme-linked immunosorbent assay (ELISA)

Tumor and liver tissues were accurately weighed, and 0.9% sterile saline was added at a ratio of 1:9 (w/v) to prepare a 10% tissue homogenate using mechanical homogenization on ice. The homogenates were centrifuged at 2500 rpm for 10 min at 4 °C, and the resulting supernatants were collected for cytokine quantification. Levels of interleukin-2 (IL-2), interleukin-6 (IL-6), interleukin-10 (IL-10), and tumor necrosis factor-alpha (TNF-α) were measured using commercially available ELISA kits according to the manufacturer’s protocols (Servicebio, Wuhan, China). Cytokine concentrations were used to assess local and systemic immune responses following treatment.

Multiplex immunofluorescence staining

Formalin-fixed, paraffin-embedded tissue sections were first dewaxed and rehydrated, then fixed in methanol for 30 min and washed three times with PBS. Endogenous peroxidase activity was blocked using 3% hydrogen peroxide (H₂O₂) for 25 min at room temperature, followed by PBS washes. After serum blocking for 30 min, the first primary antibody was applied and incubated overnight at 4 °C. Following PBS rinsing, an HRP-conjugated secondary antibody was added and incubated for 50 min at room temperature. The corresponding tyramide signal amplification (TSA) reagent was then applied for 10 min. Sections were washed with TBST and subjected to microwave treatment for 10 min to strip bound antibodies.

The above steps were repeated for the second and third rounds of staining. After sequential labeling with three different antibodies, nuclei were counterstained with DAPI for 10 min. Slides were then rinsed, incubated with an autofluorescence quenching reagent for 5 min, and coverslipped. Fluorescence images were acquired using a Nikon Eclipse C1 fluorescence microscope (Nikon, Japan). Quantitative analysis of positive cell populations, fluorescence area, and co-localization was performed using the HALO imaging analysis platform (Indica Labs, USA). The primary antibodies and TSA fluorophores used included CD4/iF555-Tyramide, CD8/iF488-Tyramide, and CD206/iF647-Tyramide (Servicebio, Wuhan, China).

In vivo tumor targeting evaluation

To assess in vivo targeting capability, SeNPs@CBT was labeled with the near-infrared fluorescent dye DiR and administered to HCT116 tumor-bearing nude mice via tail vein injection at a dose of 5 mg/kg (n = 5 per group). Free DiR was used as a control. Whole-body fluorescence imaging was performed using an IVIS Spectrum 2 system (Revvity, Inc., China) at 1, 4, 12, and 24 h post-injection. At 24 hours, mice were euthanized, and major organs (heart, liver, spleen, lung, kidney) along with tumor tissues were excised for ex vivo imaging. Quantification of fluorescence intensity and total radiant efficiency was conducted to evaluate biodistribution and tumor accumulation efficiency of the nanodrug.

In vivo biological safety evaluation

To evaluate the in vivo biosafety of SeNPs@CBT, female KM mice (5-6 weeks old) were intravenously injected via the tail vein with either PBS or SeNPs@CBT (5 mg/kg). Mice were euthanized at three designated time points: after three injections, after six injections, and several days post-final dose. Major organs, including the heart, liver, spleen, lungs, and kidneys, were collected for histopathological examination via hematoxylin and eosin (H&E) staining. In parallel, blood samples were harvested for routine hematological analysis using an automated biochemical analyzer (Chemray 800, Rayto, China).

Hemolysis assay

Fresh whole blood from KM mice was collected in anticoagulant-coated tubes and centrifuged at 1000 rpm for 10 min at 4 °C to separate the plasma. The red blood cell (RBC) pellet was washed repeatedly with PBS until the supernatant became clear. A 10% RBC suspension was then prepared in PBS. For hemolysis evaluation, 500 μL of the RBC suspension was mixed with 500 μL of either SeNPs@BT or SeNPs@CBT at concentrations of 5, 10, or 20 μg/mL and incubated at 37 °C for 4 hours. After incubation, the samples were centrifuged (1000 rpm, 10 min), and the absorbance of the supernatant was measured at 542 nm. PBS and ddH₂O were used as negative and positive controls, respectively. The hemolysis rate was calculated using the formula:

All experiments were conducted in triplicate.

Western blotting

HCT116 cells were seeded into 6-well plates at a density of 1 × 10⁶ cells per well and incubated for 24 h with SeNPs@CBT. Total proteins were extracted using RIPA lysis buffer supplemented with PMSF, followed by denaturation in boiling water for 10 min. Equal amounts of protein were separated by 15% SDS-PAGE and transferred onto PVDF membranes. After blocking with 5% skim milk for 2 hours, membranes were incubated sequentially with primary antibodies against p53, CDK1, PLK1, CYCS, and NDUFS3, followed by corresponding secondary antibodies. β-Actin was used as a loading control. Protein bands were visualized using the TanON-4600SF imaging system (Tanon Science & Technology Co., Ltd., China). All antibodies were purchased from PTM BIO (Hangzhou, China).

Quantitative reverse transcription PCR (RT-qPCR)

HCT116 cells (1 × 10⁶ per well) were seeded in 6-well plates and treated with SeNPs@CBT for 7 hours. Total RNA was extracted using Trizol reagent, following standard chloroform–isopropanol extraction. Briefly, after cell lysis on ice for 20 min, 200 μL of chloroform was added, followed by centrifugation at 12,000 rpm for 10 min at 4 °C. The aqueous phase was mixed with isopropanol, centrifuged again, washed with 75% ethanol, and air-dried. RNA was then dissolved in DEPC-treated water and quantified spectrophotometrically. cDNA synthesis and amplification were performed using the SYBR Green One-Step qRT-PCR Kit (Beyotime Biotechnology, Shanghai, China) under the following cycling conditions: reverse transcription at 50 °C for 30 min; initial denaturation at 95 °C for 2 min; followed by 40 cycles of denaturation at 95 °C for 5 s and annealing/extension at 60 °C for 30 s. Relative gene expression levels were calculated using the ΔΔCt method on a BIO-RAD CFX96 Touch™ Real-Time PCR Detection System. Each reaction was performed in triplicate. Primer sequences are provided in the Supplementary Information.

Statistical analysis

All experiments were conducted in triplicate or more. Data are presented as mean ± standard deviation (SD), unless otherwise indicated. Statistical comparisons between two groups were performed using an unpaired two-tailed Student’s t-test, while one-way analysis of variance (ANOVA) was used for multi-group comparisons. Statistical significance is denoted as follows: *p < 0.05; **p < 0.01; ***p < 0.001; n.s., not significant.

Results and discussion

Construction and characterization of a pH-responsive SeNP-based nanoassembly

To develop a multifunctional nanocarrier with enhanced antibacterial and antitumor capabilities, we constructed a selenium nanoparticle-based drug delivery system, SeNPs@CBT, capable of tumor targeting and pH-responsive drug release. The core component, CPBA-CTS (hereafter referred to as BT), was synthesized by covalently grafting CPBA onto CTS via amide bond formation [41, 42]. The successful conjugation and the optimal molar feeding ratio (CPBA:CTS = 1:0.5) were confirmed by ¹H NMR and FTIR spectroscopy (Figures S1, S2). SeNPs were synthesized using sodium selenite as the selenium source and L-ascorbic acid as the reducing agent, in the presence of BT, which functioned not only as a stabilizing agent but also as a morphology-directing agent. With increasing BT concentration, the particle morphology shifted from irregular, needle-like structures to uniform spherical nanoparticles, with morphological saturation achieved at BT concentration of 1 mg/mL or higher (Figure S3). The SeNPs@BT formulations prepared under optimal BT concentrations exhibited excellent colloidal stability with no visible aggregation for at least 24 h and stable optical absorbance up to 96 h (Figures S4, S5). To endow the system with pH-responsiveness and antibacterial functionality, the phenylboronic acid groups on the surface of SeNPs@BT were used to anchor caffeic acid phenethyl ester (CAPE) through reversible borate ester bonds under mildly alkaline conditions, yielding SeNPs@CBT. The CAPE loading capacity reached 350.8 μg/mg (Figure S6). After CAPE loading, approximately 0.16 μmol/mg of free CAPE remained on the surface of SeNPs@CBT, which provides a basis for subsequent targeting of sialic acid receptors on cancer cell membranes (Figure S7).

Dynamic light scattering (DLS) revealed hydrodynamic diameters of 101.3 nm for SeNPs@BT and 113.7 nm for SeNPs@CBT, with negligible size variation after 120 days of storage (117.5 nm and 132.8 nm, respectively), and polydispersity indices (PDI) consistently below 0.1, indicating excellent monodispersity (Fig. 2a). Zeta potentials remained stable at +17.6 mV and +12.9 mV, respectively (Fig. 2b). SEM and TEM imaging confirmed the spherical morphology and narrow size distribution (≈100 nm) of both nanostructures (Fig. 2c). A distinct ~3 nm surface coating was visible on SeNPs@CBT in TEM, likely corresponding to CAPE-polysaccharide conjugation via borate ester formation. X-ray photoelectron spectroscopy (XPS) further validated the successful fabrication of SeNPs@CBT. The survey spectrum showed characteristic peaks of O 1s, C 1s, and Se 3d (Fig. 2d). High-resolution Se 3d spectra displayed binding energies at ~55.2 eV and 56.2 eV, corresponding to Se 3d₅/₂ and Se 3d₃/₂, respectively, confirming the presence of elemental selenium (Se⁰) and indicating low oxidation (Fig. 2e). The B 1s spectrum exhibited a prominent peak at ~192.0 eV, indicative of B–O bonding and confirming the presence of boronic ester groups (Fig. 2f). FTIR spectra provided additional structural confirmation (Fig. 2g). Bare SeNPs showed no characteristic organic peaks, consistent with their inorganic composition. The presence of CPBA and CTS in SeNPs@BT was evidenced by broad N–H and O–H stretches (3500–3200 cm⁻¹) and C=O/C=C vibrations (1600–1500 cm⁻¹). In SeNPs@CBT, the emergence of a sharp absorption near 1700 cm⁻¹ (C=O stretching) and an aromatic ring peak at 1600 cm⁻¹ confirmed the successful incorporation of CAPE.

Fig. 2.

Characterization and pH-responsive drug release behavior of SeNPs@CBT. (a) Hydrodynamic diameter of SeNPs@BT and SeNPs@CBT measured by dynamic light scattering. (b) Zeta potential analysis. (c) Representative SEM and TEM images illustrating the morphology and size distribution of the nanoparticles. (d) XPS survey spectrum of SeNPs@CBT. (e) High-resolution XPS spectrum of Se 3d. (f) High-resolution XPS spectrum of B 1s. (g) FTIR spectra of BT, CAPE, SeNPs, SeNPs@BT, and SeNPs@CBT. (h) TEM image and hydrodynamic diameter of SeNPs@CBT after incubation at tumor-mimicking acidic pH. (i) Cumulative release profile of CAPE from SeNPs@CBT under different pH conditions

To evaluate the pH-responsive behavior of SeNPs@CBT, we investigated their morphological evolution and drug release profile under acidic conditions mimicking the TME (pH 5.0). TEM analysis revealed that SeNPs@CBT retained their spherical morphology after 12 h of incubation at pH 5.0, with a hydrodynamic diameter of approximately 109.3 nm (Fig. 2h). Notably, the acidic environment induced the detachment of the CAPE-containing surface coating, resulting in a morphology similar to SeNPs@BT, consistent with pH-triggered borate ester cleavage. The release kinetics of CAPE were further quantified (Fig. 2i). At physiological pH 7.4, cumulative CAPE release from SeNPs@CBT was limited to ~21.8% over 72 hours, whereas under pH 5.0, ~84.7% of CAPE was released within 24 hours, reaching a plateau thereafter. These findings confirm the high sensitivity of the borate ester bonds to acidic environments, enabling selective and efficient drug release specifically within the tumor milieu. Collectively, these results demonstrate that SeNPs@CBT combines structural stability under physiological conditions with pH-responsive drug release, an essential feature for precise tumor-targeted therapy.

In vitro antibacterial activity and mechanism of SeNPs@CBT against F. nucleatum

The antibacterial efficacy of SeNPs@CBT against F. nucleatum was first evaluated using a standard colony-forming assay. As shown in Fig. 3a and 3b, SeNPs@BT alone exhibited limited antibacterial activity, with a bacterial survival rate of 61.9% even at 25 μg/mL. Free CAPE, administered at 20 μg/mL, inhibited approximately 79.6% of bacterial growth, but its efficacy was constrained by the structural defenses of Gram-negative bacteria, particularly the lipopolysaccharide-rich outer membrane that impedes drug penetration, as well as CAPE’s poor solubility and bioavailability [43]. In contrast, SeNPs@CBT achieved a marked reduction in bacterial survival to 6.8% at a lower CAPE-equivalent concentration of 10 μg/mL, demonstrating a substantial enhancement in antibacterial potency. This improvement likely stems from the nanoparticle-mediated facilitation of CAPE delivery across bacterial membranes and its localized enrichment at the bacterial surface.

Fig. 3.

In vitro antibacterial activity and underlying mechanisms of SeNPs@CBT against F. nucleatum. (a) Representative colony formation images. (b) Quantitative analysis of F. nucleatum viability after treatment with various formulations. (c) Inhibitory effects of SeNPs@CBT and CAPE on both intracellular and extracellular F. nucleatum in HCT116 cells. (d) SEM images illustrating morphological alterations of F. nucleatum following different treatments. (e) Live/dead fluorescence staining of F. nucleatum post-treatment (Scale Bar = 20 μm). (f) Intracellular ROS levels in F. nucleatum induced by drug exposure (Scale Bar = 30 μm). (g) Impact of various treatments on the membrane potential of F. nucleatum (Scale Bar = 50 μm)

We further assessed the capacity of SeNPs@CBT to eradicate both extracellular and intracellular F. nucleatum using an HCT116 CRC cell infection model (Fig. 3c, Figure S8). At the same CAPE concentration (10 μg/mL), the survival rate of extracellular bacteria in the cellular context was slightly elevated to 36.2%, compared to the in vitro condition, yet SeNPs@CBT consistently outperformed free CAPE in eliminating intracellular bacteria. This enhanced intracellular efficacy may be attributed to the Warburg effect of tumor cells, where heightened glycolysis acidifies the extracellular environment, partially triggering CAPE release from SeNPs@CBT prior to cellular uptake [44]. This process exposes phenylboronic acid groups on the nanoparticle surface, promoting selective binding to sialic acid receptors overexpressed on HCT116 cells and facilitating targeted endocytosis. Consequently, SeNPs@CBT achieved superior intracellular bacterial clearance across all tested concentrations.

Scanning electron microscopy (SEM) provided visual confirmation of antibacterial effects on bacterial morphology (Fig. 3d). Untreated F. nucleatum exhibited typical spindle-shaped structures with well-defined boundaries. In contrast, bacteria exposed to CAPE or SeNPs@CBT showed pronounced morphological disruptions, including cell wall damage, distorted shapes, and blurred boundaries, indicative of membrane compromise. SeNPs@BT treatment resulted in nanoparticle adsorption onto bacterial surfaces without inducing significant structural damage. Bacterial viability was further quantified using a live/dead fluorescence staining assay (Fig. 3e, Figure S9). DMAO stains all bacteria (green fluorescence), while propidium iodide (PI) penetrates only compromised membranes, marking dead bacteria with red fluorescence [45]. At 15 μg/mL, free CAPE predominantly yielded green fluorescence within the first 3 hours, with substantial viable bacteria persisting even at 5 hours. In contrast, SeNPs@CBT treatment induced pronounced yellow fluorescence as early as 3 hours, indicating rapid and effective bacterial killing that intensified over time. To elucidate the antibacterial mechanism, reactive oxygen species (ROS) generation in F. nucleatum was assessed using a ROS-sensitive fluorescent probe (Fig. 3f). Minimal basal ROS levels were detected in the PBS control group. Free CAPE induced limited ROS accumulation, whereas SeNPs@CBT triggered a robust and uniformly distributed green fluorescence signal, indicating significant ROS elevation. This suggests that SeNPs@CBT synergistically amplifies CAPE’s oxidative stress-inducing capacity, contributing to bacterial killing via oxidative damage. Additionally, bacterial membrane potential disruption was evaluated using the fluorescent dye DiOC2(3) (Fig. 3g, Figure S10). In healthy bacteria with intact membrane potential, DiOC2(3) aggregates, emitting red fluorescence. Membrane depolarization, induced by agents such as the positive control CCCP (carbonyl cyanide 3-chlorophenylhydrazone), shifts the emission towards green [46, 47]. Following SeNPs@CBT treatment, a significant reduction in red fluorescence was observed, indicative of membrane depolarization comparable to CCCP treatment. In contrast, free CAPE caused only modest membrane potential changes.

In summary, SeNPs@CBT exhibits potent antibacterial activity against F. nucleatum through a dual mechanism involving pronounced ROS generation and disruption of bacterial membrane potential. This multi-targeted antibacterial strategy not only enhances bactericidal efficacy but may also reduce the likelihood of resistance development. The ROS-mediated, nonspecific oxidative damage pathway presents a valuable approach to overcoming the inherent defenses of Gram-negative bacteria and offers a promising alternative to conventional antibiotics.

Cytotoxicity, microenvironment adaptability, and targeted uptake of SeNPs@CBT

Following the demonstration of SeNPs@CBT’s potent antibacterial effects, we systematically evaluated its cytotoxicity, biological safety, microenvironment adaptability, and tumor-targeting ability in vitro. The cytotoxicity of CAPE, SeNPs@CTS, SeNPs@BT, and SeNPs@CBT was assessed in CRC cells and normal cell lines via the MTT assay. CAPE exhibited dose-dependent inhibitory effects on two CRC cell lines, HCT116 and HT-29. However, at 40 μg/mL (0.141 mM), CAPE also induced significant cytotoxicity in normal hepatocytes (HL7702), reducing viability below 50%, whereas CRC cells still exhibited >40% viability at this concentration (Fig. 4a). These findings suggest that although CAPE possesses anticancer potential, its intrinsic selectivity toward malignant cells over normal cells is suboptimal. In contrast, SeNPs@CBT exhibited pronounced concentration- and time-dependent cytotoxicity in HCT116 and HT-29 cells, with an IC50 of approximately 6 μg/mL, which was significantly lower than that of SeNPs@BT (IC50 >10 μg/mL) (Fig. 4b). SeNPs@CTS exhibited markedly lower cytotoxicity toward both cell lines than SeNPs@BT and SeNPs@CBT, with cell viability remaining above 75% at 10 μg/mL (Figure S11a). This observation indicates that, in the absence of CPBA grafting, selenium nanoparticles lack effective tumor-targeting capability, thereby requiring higher concentrations to achieve significant inhibitory effects. Accordingly, subsequent studies and mechanistic investigations were primarily focused on the SeNPs@BT and SeNPs@CBT systems, which exhibit greater therapeutic potential. After 72 h of treatment, the viability of both CRC cell lines in the SeNPs@CBT group decreased to below 20%, indicating potent antitumor efficacy (Fig. 4c). Importantly, SeNPs@BT showed negligible toxicity toward normal HL7702 and HCoEpiC cells, with survival rates maintained above 65% even at 100 μg/mL. Although SeNPs@CBT exhibited slightly elevated toxicity toward normal cells, likely due to CAPE incorporation, cell viability remained above 90% at 10 μg/mL (Figure S11b,c), affirming favorable biological safety within the therapeutic dosage range.

Fig. 4.

Evaluation of cytotoxicity and cellular uptake of SeNPs@CBT in cancer cells. (a) Inhibitory effects of CAPE on the proliferation of normal cells (HL7702, HCoEpiC) and CRC cells (HCT116, HT-29). (b) Cytotoxic effects of SeNPs@BT and SeNPs@CBT on HCT116 and HT-29 cells. (c) Time-dependent viability of HCT116 and HT-29 cells following SeNPs@CBT treatment. (d) Inhibition of F. nucleatum-enhanced proliferation in HCT116 cells by SeNPs@BT and SeNPs@CBT. (e) Cellular uptake efficiency of SeNPs@CBT across different cell lines. (f) Time-dependent intracellular localization of SeNPs@CBT in HCT116 cells visualized by confocal microscopy (Scale Bar = 100 μm)

To further simulate the pathological conditions of CRC, we examined the impact of F. nucleatum infection on CRC cell proliferation and drug efficacy. Infection with F. nucleatum promoted HCT116 proliferation by approximately 21.9%, while significantly attenuating the cytotoxic effects of SeNPs@BT. In contrast, SeNPs@CBT retained stable, dose-dependent cytotoxicity under these conditions, with only a modest increase in IC50 to ~8 μg/mL (Fig. 4d). This underscores the resilience of SeNPs@CBT’s antitumor activity even within a bacteria-enriched TME, an essential feature for clinical relevance. Subsequently, we investigated the tumor-targeting capability of SeNPs@CBT. Sialic acid (SA), commonly overexpressed on tumor cell surfaces, facilitates immune evasion, adhesion, and metastasis, making it an attractive molecular target [48]. The phenylboronic acid moieties on SeNPs@CBT enable specific recognition of SA residues, thereby promoting tumor-selective binding. Competitive experiments showed that treatment of HCT116 cells with 1 μM CAPE for 15 min markedly reduced the cytotoxicity of SeNPs@CBT, with cell viability exceeding 56% at a concentration of 10 μg/mL (Figure S12). Cellular uptake studies across six tumor cell lines, including HCT116 and HT-29, demonstrated significant internalization of SeNPs@CBT, whereas uptake was markedly reduced in normal cells such as HL7702 and HCoEpiC, which exhibit low SA expression (Fig. 4e). Further validation via confocal microscopy in Dil-labeled SeNPs@CBT-treated HCT116 cells revealed progressively increasing intracellular red fluorescence over time, with pronounced perinuclear aggregation (Fig. 4f), indicating efficient cellular penetration and accumulation.

To elucidate the underlying mechanism, we propose a plausible endocytic pathway. Tumor cells often display elevated levels of SA on their plasma membranes [48]. Phenylboronic acid moieties on SeNPs@CBT can reversibly form boronate ester bonds with the cis-diol groups of SA residues [49], enabling selective recognition and anchoring. This interaction may promote clustering of SA-bearing glycoproteins or receptors, stabilizing nanoparticle adhesion and prolonging membrane residence time [50]. As SA itself lacks intrinsic signaling capability, internalization of the bound nanoparticles is likely mediated by secondary endocytic processes. Previous studies have indicated that clathrin-dependent uptake, potentially involving EGFR/PI3K pathway activation, may facilitate such internalization [51]. Following endocytosis, PBA-functionalized nanoparticles preferentially localize to lysosomes [52], where acidic and enzyme-rich conditions promote degradation of the chitosan shell and subsequent drug release [53]. This lysosomal activation enhances intracellular selenium accumulation and cytotoxic potency. Though this mechanism remains to be further validated, it provides a reasonable working hypothesis to account for the observed tumor-targeting and uptake behavior of SeNPs@CBT.

Collectively, these results demonstrate that SeNPs@CBT possesses robust antitumor activity against CRC cells, maintains high biocompatibility at therapeutic concentrations, and effectively targets tumor cells via sialic acid-mediated recognition. Furthermore, its sustained cytotoxic efficacy in the presence of F. nucleatum highlights its adaptability within complex TMEs. These attributes provide a solid mechanistic and translational foundation for the application of SeNPs@CBT in treating CRC subtypes characterized by high microbial burden, as well as other SA-overexpressing malignancies.

In vivo antitumor efficacy of SeNPs@CBT in a CRC mouse model colonized by F. nucleatum

Given the promising in vitro cytotoxicity and targeting capability of SeNPs@CBT, we next evaluated it’s in vivo antitumor efficacy using a CRC mouse model established through infection with F. nucleatum. A subcutaneous tumor model was established by implanting HCT116 cells into the axillary region of BALB/c nude mice. Once the tumors reached approximately 200 mm³, F. nucleatum was intratumorally injected at multiple sites to simulate a bacteria-enriched TME, following established protocols [54, 55]. The experimental groups received intravenous administrations of PBS, SeNPs@BT, or SeNPs@CBT (5 mg/kg) every other day for a total of six doses (Fig. 5a). Tumor volume and body weight were monitored throughout the treatment period (Fig. 5b-d), and tumors were harvested and weighed on day 14 (Fig. 5e, f).

Fig. 5.

In vivo antitumor efficacy of SeNPs@CBT in a subcutaneous HCT116 tumor model colonized by F. nucleatum. (a) Schematic of the treatment regimen in tumor-bearing mice (n=6 per group). (b) Average tumor volume progression and (c) Individual tumor growth trajectories across different treatment groups. (d) Changes in body weight during the treatment period. (e) Representative images of excised tumors from each group and (f) Corresponding tumor weights at the endpoint. (g) TUNEL staining of tumor sections to evaluate apoptosis induction post-treatment (Scale Bar, 200 μm). (h) Quantification of apoptotic cells based on integrated fluorescence intensity

F. nucleatum infection significantly accelerated tumor growth in the PBS (+) group, resulting in pronounced increases in both tumor volume and mass compared to uninfected controls. These results are consistent with our in vitro findings and previous studies [56–59]. In the uninfected SeNPs@BT-treated group, tumor growth was inhibited by approximately 46%, but this inhibition was reduced to 34.6% in the presence of F. nucleatum, indicating that microbial interference compromises therapeutic efficacy. Remarkably, SeNPs@CBT exhibited robust antitumor activity even under microbial challenge, achieving up to 96% tumor growth inhibition, with complete tumor eradication observed in some mice (Fig. 5b-f). Importantly, SeNPs@CBT treatment was well tolerated, as evidenced by stable body weights comparable to the healthy control group throughout the treatment period. In contrast, mice in the PBS (+) group exhibited an average body weight loss of 7.67%, reflecting the increased tumor burden. To further investigate the therapeutic mechanism, TUNEL staining was performed to assess apoptosis induction in tumor tissues (Fig. 5g, h). Minimal apoptosis (<5% TUNEL-positive cells) was detected in the PBS and PBS (+) groups. SeNPs@BT and SeNPs@BT (+) induced moderate apoptosis rates of 34.2% and 23.8%, respectively. In striking contrast, the SeNPs@CBT (+) group exhibited widespread apoptosis, with over 75% of tumor cells TUNEL-positive, nearly saturating the entire tumor section with green fluorescence.

Collectively, these in vivo results demonstrate that SeNPs@CBT not only exhibits potent antitumor activity but also effectively mitigates the tumor-promoting influence of F. nucleatum infection. The superior therapeutic performance of SeNPs@CBT is attributable to its dual-action mechanism: the CAPE-mediated antibacterial effect counteracts the pro-tumorigenic impact of bacterial colonization, while the intrinsic cytotoxicity of the selenium core induces pronounced tumor cell apoptosis. These findings establish SeNPs@CBT as a promising candidate for treating microbiota-associated CRC and potentially other malignancies characterized by a dysbiotic TME.

SeNPs@CBT modulates the tumor-associated immune microenvironment through therapeutic intervention

Given the pronounced in vivo antitumor efficacy of SeNPs@CBT, we further explored whether its therapeutic benefits extend to the modulation of the tumor-associated immune microenvironment, particularly in the context of F. nucleatum colonization. Selenium, as an essential trace element, is well-known for its immunoregulatory functions in the human body [60–63]. To this end, we evaluated the immune status within tumor and liver tissues following treatment across different experimental groups.

Cytokines play pivotal roles in orchestrating host immune responses against cancer and critically influence tumor progression and therapeutic outcomes [64, 65]. As shown in Fig. 6a, we quantified key cytokines, including IL-10, IL-6, IL-2, and TNF-α, within tumor tissues. IL-10, a potent immunosuppressive cytokine that limits excessive inflammation and tissue damage [66], was markedly elevated in the PBS and PBS (+) groups, likely due to the LPS-mediated inflammatory response induced by F. nucleatum [67]. This elevation persisted in the SeNPs@BT and SeNPs@BT (+) groups; however, IL-10 levels were significantly reduced in the SeNPs@CBT (+) group, suggesting that CAPE’s anti-inflammatory and immunomodulatory effects contributed to restoring immune balance. IL-6, a pleiotropic pro-inflammatory cytokine that reflects immune system activation and plays complex roles in both pro- and anti-tumor responses [68], was elevated in both SeNPs@BT and SeNPs@CBT treatment groups, suggestive of acute local immune activation in response to treatment-induced tumor stress. Notably, IL-6 levels in the SeNPs@CBT (+) group were relatively lower compared to the other treatment groups. Given the predominantly immunosuppressive and tumor-promoting effects of chronic IL-6 elevat7ion in the tumor microenvironment (e.g., via STAT3 activation and myeloid-derived suppressor cell recruitment), this reduction likely reflects diminished tumor burden—as tumors and associated stromal cells are major sources of persistent IL-6 production—along with attenuated chronic inflammation and potentially improved antitumor immune responses, consistent with prior reports linking IL-6 normalization to effective tumor control and better outcomes [69]. IL-2, essential for T cell activation and proliferation [70], was significantly upregulated in both SeNPs@BT and SeNPs@CBT groups, highlighting the nanocarriers’ ability to stimulate T cell-mediated immunity. Furthermore, TNF-α levels, indicative of tumor antigen-specific immune responses [71], were consistently increased, further corroborating the activation of antitumor immunity. In addition to the tumor site, the liver serves as a central immunological organ and a critical indicator of systemic immune modulation [72]. Cytokine expression patterns in the liver (Fig. 6b) were consistent with those observed in tumor tissues, suggesting that SeNPs@CBT exerts bidirectional immunoregulatory effects by enhancing both local and systemic immune environments.

Fig. 6.

Immunomodulatory effects of SeNPs@CBT on the TME. Cytokine expression profiles in tumor tissues (a) and liver tissues (b) from mice across different treatment groups. (c) Immunofluorescence staining of tumor sections showing immune cell populations post-treatment (Scale Bar = 1 mm). (d) Quantitative analysis of immune cell subpopulations within the TME

To further delineate immune cell dynamics, we analyzed immune cell subpopulations within tumor tissues (Fig. 6c, d). SeNPs@BT treatment significantly increased the proportions of CD8⁺ cytotoxic T cells and CD4⁺ memory T cells, indicating effective recruitment and activation of antitumor lymphocytes. However, this effect was attenuated in the SeNPs@BT (+) group, demonstrating that F. nucleatum colonization can suppress immune activation. Interestingly, in the SeNPs@CBT (+) group, the final proportion of T cells was relatively lower, which may reflect the resolution of inflammation and restoration of immune homeostasis following extensive tumor regression and apoptosis. A particularly notable finding was the significant reduction of M2 macrophages in the SeNPs@CBT (+) group. M2 macrophages, often associated with immunosuppression and tumor progression via collaboration with myeloid-derived suppressor cells (MDSCs) [73, 74], were minimized, indicating a reprogramming of the immunosuppressive TME towards a more immunostimulatory state.

In summary, SeNPs@CBT effectively eradicates F. nucleatum and concurrently remodels the tumor immune microenvironment through a dual mechanism: attenuating bacterial-driven inflammation and immunosuppression, while enhancing T cell activation and inhibiting M2 macrophage polarization. This strategy offers a compelling platform for microbiota-targeted tumor immunotherapy and establishes a conceptual framework for addressing microorganism-associated cancers.

In vivo biodistribution and biosafety assessment of SeNPs@CBT

Encouraged by the potent antitumor and immunomodulatory effects of SeNPs@CBT, we further investigated its in vivo biodistribution and biosafety to evaluate its translational potential. For real-time tracking, SeNPs@CBT was labeled with the near-infrared fluorescent probe DiR, while free DiR served as a control. Fluorescence imaging was conducted at 1, 4, 12, and 24 h post-injection in tumor-bearing mice (Fig. 7a). As expected, free DiR, due to its small molecular size, rapidly distributed through body fluids and accumulated in metabolic organs, particularly the liver and abdominal cavity, within the first hour. No significant tumor accumulation was observed even after 24 hours. In contrast, SeNPs@CBT gradually accumulated at the tumor site beginning at 4 h post-injection, a process likely mediated by the EPR effect of nanoparticles. The fluorescence signal intensified over time, with predominant localization in the subcutaneous tumor region by 24 hours. Three-dimensional simulated fluorescence imaging further visualized the tumor-specific distribution of SeNPs@CBT (Figure S13a). Ex vivo imaging of excised major organs and tumor tissues at 24 h confirmed the in vivo observations (Fig. 7b, c). While free DiR was primarily distributed in the liver, spleen, and kidneys, SeNPs@CBT exhibited strong fluorescence selectively at the tumor site, with minimal signals in other organs. These results underscore the favorable tumor-targeting ability and tissue specificity of SeNPs@CBT.

Fig. 7.

Biodistribution and in vivo safety evaluation of SeNPs@CBT. (a) Time-dependent biodistribution of SeNPs@CBT in tumor-bearing mice following tail vein injection; the tumor site is indicated by the blue ellipse in the 1 h image. (b) Ex vivo fluorescence imaging of major organs and tumor tissues. (c) Quantitative analysis of total fluorescence intensity 24 h post-injection. (d) Schematic diagram of the in vivo safety assessment protocol for SeNPs@CBT. (e) Body weight changes in mice across different treatment groups. (f) Hematological parameters reflecting systemic toxicity. (g) Hemolysis rate of erythrocytes following SeNPs@CBT exposure. (h) Histopathological evaluation of major organs by H&E staining (Scale Bar = 0.1 mm)

To comprehensively evaluate biosafety, we conducted a systemic toxicity assessment in healthy mice across three time points: after three injections, after six injections, and several days following the final dose. Throughout the 20-day observation period, all mice maintained stable body weights and normal behavior, with no statistically significant differences compared to controls (Fig. 7d, e). Routine blood analyses revealed that red blood cells (RBC), hemoglobin (HGB), and platelets (PLT) remained within physiological ranges, indicating no hemolytic effects or coagulation dysfunction (Figure S13b). White blood cell (WBC), neutrophil (NEUT), and lymphocyte (LYM) counts showed no abnormal fluctuations, and eosinophil (EOS) and basophil (BASO) levels remained stable, suggesting an absence of inflammatory responses, allergic reactions, or immune disturbances. The hemolytic potential of SeNPs@CBT was further evaluated in vitro, where the hemolysis rate remained below 1% at concentrations up to 20 μg/mL, well below the accepted safety threshold of 5% (Fig. 7g), confirming excellent blood compatibility. Histopathological analysis of major organs, including the heart, liver, spleen, lungs, and kidneys, via hematoxylin and eosin (H&E) staining, revealed no significant pathological changes or tissue damage (Fig. 7g, Figure S13c). The structural integrity of hepatic central veins and renal glomeruli was preserved, and no observable differences were detected between the SeNPs@CBT-treated and PBS control groups.

Collectively, these findings demonstrate that SeNPs@CBT possesses not only effective tumor-targeting capability but also excellent systemic biosafety, supporting its potential for clinical translation in microbiota-associated cancer therapy. This comprehensive safety profile, alongside its antitumor and immunomodulatory efficacy, underscores SeNPs@CBT as a promising nanotherapeutic platform for precision oncology.

Transcriptomic analysis of HCT116 cell apoptosis induced by SeNPs@CBT

To further elucidate the molecular mechanisms underlying SeNPs@CBT-induced apoptosis in CRC cells, we performed transcriptomic profiling of HCT116 cells following SeNPs@CBT treatment. Differential gene expression analysis identified 2,923 upregulated and 1,599 downregulated genes with statistically significant changes compared to untreated controls (Fig. 8a, b). Gene Ontology (GO) enrichment analysis revealed that these differentially expressed genes (DEGs) were predominantly involved in biological processes such as protein binding, DNA binding, cell cycle regulation, mitotic division, and transcriptional control (Fig. 8c). Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis further indicated that the p53 signaling pathway, cell cycle pathway, protein processing in the endoplasmic reticulum, and several metabolism-related pathways were significantly enriched (Fig. 8d), suggesting that SeNPs@CBT orchestrates apoptosis through a multifaceted regulatory network.

Fig. 8.

Mechanistic investigation of SeNPs@CBT-induced apoptosis in HCT116 cells. (a) Statistical summary of differentially expressed genes identified by transcriptomic analysis. (b) Volcano plot of gene expression changes following SeNPs@CBT treatment. (c) KEGG pathway enrichment analysis displaying the top 20 pathways with the smallest p-values. (d) GO enrichment analysis of biological processes, showing the top 20 significantly enriched terms. (e) RT-qPCR validation of gene expression changes; the suffix C denotes the control group and T denotes the SeNPs@CBT-treated group. (f) Western blot analysis of protein expression levels in HCT116 cells after SeNPs@CBT treatment. (g) Flow cytometric analysis of cell cycle distribution in HCT116 cells after 12 h of SeNPs@CBT exposure

To validate key apoptotic and cell cycle regulatory genes identified in the transcriptomic analysis, representative markers—including CYCS, NDUFS3, TP53, MDM2, PLK1, and CDK1, were selected for RT-qPCR quantification (Fig. 8e), and protein expression levels were assessed by Western blotting (Fig. 8f). CYCS (cytochrome c) was notably upregulated, indicating mitochondrial outer membrane permeabilization and the subsequent cytoplasmic release of cytochrome c, a hallmark of intrinsic apoptosis pathway activation. NDUFS3, a critical component of mitochondrial complex I, was downregulated, implying impaired electron transport chain function, reduced ATP production, and potential ROS accumulation. The tumor suppressor p53 was significantly upregulated, consistent with activation of p53-mediated apoptotic signaling and cell cycle arrest. Interestingly, MDM2, a negative feedback regulator of p53, was concurrently upregulated, potentially reflecting a cellular attempt to mitigate p53 activation during stress responses. The downregulation of PLK1 and CDK1 suggests inhibition of cell cycle progression, particularly at key checkpoints controlling mitotic entry and transition.

These molecular alterations translated into phenotypic cell cycle effects, as flow cytometry revealed significant G1 phase arrest in SeNPs@CBT-treated HCT116 cells, with concomitant reductions in S and G2/M phase populations (Fig. 8g). This suggests that SeNPs@CBT impairs the G1/S transition, potentially through inhibition of the CDK2-Cyclin E complex, aligned with p53 activation and CDK1/PLK1 suppression. Furthermore, mitochondrial dysfunction resulting from NDUFS3 downregulation likely exacerbates cell cycle arrest by limiting bioenergetic capacity required for DNA synthesis and mitosis.

Together, these findings demonstrate that SeNPs@CBT induces HCT116 cell apoptosis via a dual mechanism involving mitochondrial dysfunction and p53-driven cell cycle arrest, supported by coordinated transcriptional and translational regulation. This mechanistic insight complements the observed antitumor efficacy in vivo and further establishes the potential of SeNPs@CBT as a multifunctional therapeutic platform (Fig. 9).

Fig. 9.

Schematic illustration of SeNPs@CBT-induced DNA damage and mitochondrial dysfunction. SeNPs@CBT trigger DNA damage, thereby activating the ATM/ATR–p53 signaling pathway, which inhibits the activity of Cdc25 and Cyclin–CDK complexes, leading to G2/M phase arrest. Meanwhile, SeNPs@CBT impair the mitochondrial electron transport chain and attenuate oxidative phosphorylation, collectively promoting cell death. (Red upward arrows indicate significant upregulation, and green downward arrows indicate significant downregulation.)

Conclusion

In summary, we developed a multifunctional nanomedicine platform, SeNPs@CBT, integrating antibacterial, antitumor, and tumor-targeting properties for the effective treatment of CRC associated with F. nucleatum infection. In the acidic microenvironment of tumors, the borate ester bonds undergo pH-triggered cleavage, enabling the controlled release of CAPE to eradicate F. nucleatum and mitigate its pro-tumorigenic effects. Simultaneously, the exposure of additional phenylboronic acid groups enhances the tumor-targeting capability of SeNPs@CBT, promotes cellular uptake, and improves the efficiency of intracellular bacterial clearance. Compared to CAPE alone, SeNPs@CBT exhibited more pronounced efficacy in eliminating intracellular F. nucleatum. Importantly, SeNPs@CBT demonstrated excellent biosafety profiles, capable of activating anti-tumor immune responses, promoting remodeling of the tumor immune microenvironment, and enhancing the immune system’s capacity to clear tumor cells. Transcriptome analysis further revealed that SeNPs@CBT can exert synergistic anticancer effects by inducing cell cycle arrest and apoptosis through the regulation of multiple signaling pathways. Collectively, SeNPs@CBT, based on a " dual elimination of bacteria and tumors " strategy, not only achieves synergistic therapy through the precise elimination of symbiotic bacteria and targeted destruction of tumor cells, but also establishes a novel intervention model for microbiota-tumor interactions. This approach introduces a ternary therapeutic paradigm encompassing "antibacterial action, immune regulation, and direct tumor killing". This study offers a promising and feasible strategy for the development of nanotherapeutic approaches against other microbiota-associated tumors, such as gastric and pancreatic cancers. Although SeNPs@CBT exhibits favorable synergistic effects, the potential impact of selenium nanoparticles on host selenium homeostasis and the cellular internalization mechanisms warrants further investigation. In future work, advanced approaches—such as fluorescent labeling combined with confocal laser scanning microscopy, assessments using pathway-specific endocytosis inhibitors, and live-cell tracking—may be employed to more precisely elucidate its uptake behavior. Importantly, SeNPs@CBT holds broad promise for targeted therapy of colorectal cancer. We remain committed to further clarifying mechanistic details and optimizing its design and application, with particular emphasis on pharmacokinetics, biodistribution, and dose-optimization strategies to support safe and effective clinical translation.

Supporting Information

¹H NMR and FTIR spectra of BT, SEM images of SeNPs@BT modified with different BT concentrations, stability profiles of SeNPs@BT, complete CLSM images, solid agar plate photographs of intracellular and extracellular F. nucleatum, cytotoxicity toward normal cells, cytotoxicity following CAPE treatment, in vivo 3D simulated imaging of drug distribution, hematological analysis, H&E staining images of major organs, gating strategy for flow cytometry analysis, RT-qPCR primer sequences, and uncropped Western blot images.

Abbreviations

F. nucleatum

Fusobacterium nucleatum

CRC

Colorectal cancer

CAPE

Caffeic acid phenethyl ester

TME

Tumor microenvironment

EPR

Enhanced permeability and retention

SeNPs

Selenium nanoparticle

CTS

Chitosan

CPBA

4-carboxyphenylboronic acid

BT

Chitosan conjugated with 4-carboxyphenylboronic acid

SeNPs@BT

BT-modified selenium nanoparticles

SeNPs@CBT

SeNPs@BT conjugated with caffeic acid phenethyl ester

MTT

3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide

HCT116

Human colorectal carcinoma cell line

HT-29

Human colorectal adenocarcinoma cell line

KYSE-150

Human esophageal squamous cell carcinoma cell line

SW579

Human thyroid carcinoma cell line

AGS

Human gastric adenocarcinoma cell line

A375

Human malignant melanoma cell line

HL7702

Human normal liver cell line (L-02)

HcoEpiC

Human colon epithelial cell line

DMEM

Dulbecco’s modified eagle medium

RPMI

Roswell park memorial institute

FBS

Fetal bovine serum

NHS

N-hydroxysuccinimide

EDC

1-Ethyl-3-(3-dimethylaminopropyl) carbodiimide

MWCO

Membrane with a molecular weight cut-off

Na₂SeO₃

Sodium selenite

DMSO

Dimethyl sulfoxide

SEM

Scanning electron microscopy

TEM

Transmission electron microscopy

XPS

X-ray photoelectron spectroscopy

FTIR

Fourier-transform infrared

¹H NMR

Proton nuclear magnetic resonance

UV–Vis

Ultraviolet-visible

PBS

Phosphate-buffered saline

TSB

Tryptic soy broth

mTSB

Modified tryptic soy broth

CFU

Colony-forming units

ROS

Reactive oxygen species

NIH

National institutes of health

IACUC

Institutional animal care and use committee

CAPL

China animal protection law

TUNEL

Terminal deoxynucleotidyl transferase dUTP nick end labeling

ELISA

Enzyme-linked immunosorbent assay

IL-2

Interleukin-2

IL-6

Interleukin-6

IL-10

Interleukin-10

TNF-α

Tumor necrosis factor-alpha

H₂O₂

Hydrogen peroxide

TSA

Tyramide signal amplification

DAPI

4,6-Diamidino-2-phenylindole

DiR

1,1-Dioctadecyl-3,3,3,3-tetramethylindotricarbocyaineiodide

H&E

Hematoxylin and eosin

RBC

Red blood cell

p53

Tumor Protein p53

CDK1

Cyclin-dependent kinase 1

PLK1

Polo-like kinase 1

CYCS

Cytochrome c, somatic

NDUFS3

NADH:Ubiquinone oxidoreductase core subunit S3

RT-qPCR

Quantitative reverse transcription PCR

Author contributions

Chunshan Quan, Liying Zhang, Baoquan Liu: Writing – review & editing, Methodology, Investigation, Formal analysis, Conceptualization. Haodi Ma: Writing – original draft, Visualization, Investigation. Lulu Wang, Junfeng Liu, Shuai Ge, Haoyang Sun: Methodology, Investigation.All authors reviewed the manuscript

Funding

The authors acknowledge the support provided by the Applied Basic Research Project of Liaoning Provincial Science and Technology Department (2022020332-JH2/1013) and the Basic scientific research projects of Liaoning Provincial Department of Education (LJKMZ20220401).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethical approval and consent to participate

All animal experiments were conducted in compliance with the guidelines of the National Institutes of Health (NIH) and were approved by the Institutional Animal Care and Use Committee (IACUC) of Dalian Medical University.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.