Bioorthogonal catalytic centres engineered for gastrointestinal stabilization provide oral delivery for the treatment of gastric cancer

Abstract

Background

Gastric cancer, the fifth most prevalent cancer globally, poses significant treatment challenges due to factors such as late diagnosis, early metastasis, limited surgical options, and the systemic toxicity of chemotherapy. Because luminal barriers are often compromised in gastric cancers , orally administered therapies that enable localized absorption and drug release represent a promising new direction for site-specific treatment with limited side effects.

Results

We introduced disulfide-linked thermostable exoshell system that orally delivered protein-based bioorthogonal catalytic centres directly to cancer tissues. The highly engineered exoshells effectively encapsulated and stabilized labile catalytic centres, preventing degradation in the harsh gastric environment. In vivo gastric tumors were treated using the anti-cancer properties of active metabolites of the prodrug indole-3-acetic acid (IAA) converted in situ via bioorthogonal catalysis. In vitro cell studies revealed a dose- and time-dependent inhibition of gastric cancer cell growth, irrespective of their HER2 status. This inhibition was accompanied by upregulation of mitochondrial lipid peroxidation, reduced mitochondrial membrane potential, and activation of necroptotic pathway markers such as RIP1, RIP3, and MLKL at both mRNA and protein levels. In a mouse model of gastric cancer induced by N-Methyl-N-Nitrosourea, oral administration of catalytic exoshells for 6 weeks significantly inhibited gastric inflammation and tumour polyp growth. Additionally, LC/MS/MS-based metabolomic analysis of plasma obtained from treated mice showed significant upregulation of cytotoxic metabolites of IAA. Notably, metabolites relevant to redox regulation, including alpha-tocopherol (vitamin E), glutathione (GSH), homocysteine, methyl cysteine, and cysteine sulfinic acid, were identified as the top differentially expressed metabolites, indicating potent suppression of inflammation and tumour growth. Histological analysis of gastric tissue showed a reduced number of polyps and subsequent development of gastric tumours.

Conclusion

Our in vitro and in vivo results demonstrated that exoshells possessed significant potential as an orally administered, titratable therapeutic platform for the management of gastrointestinal cancers.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-026-04065-2.

Introduction

According to the Global Cancer Statistics, gastric cancer (GC) is both the fifth most prevalent cancer in the world and the fourth leading cause of death globally [1]. There are two forms of GC: intestinal (ulcerating glandular cells) and diffuse (thickening of cell layers in stomach region), with adenocarcinomas (glandular cell cancer), accounting for the majority of the cases [2]. In general, risk factors for GC include Helicobacter pylori infections, genetic mutations, consumption of alcohol and high-sodium foods, and obesity [3, 5]. GC cells overexpress surface receptors including programmed death-ligand 1 and receptor tyrosine-protein kinase erbB-2 (HER2), resulting in enhanced anti-apoptotic signalling and uninterrupted proliferation [6, 7]. Current treatments for GC include invasive approaches such as surgical resection, neoadjuvant and adjuvant chemotherapy and immunotherapy. However, because GC is more commonly diagnosed in later stages, surgical resection is limited in ensuring remission for metastasized cancers. Furthermore, the emergence of multidrug resistance and tumour heterogeneity have restricted the success of chemotherapy [8, 10]. As a result, developing precise methods for early detection and diagnosis, as well as more effective treatment alternatives remain a poignant two-pronged solution to the challenges posed by such an aggressive cancer type.

Nanoparticle-mediated oral delivery platforms have been developed with clinically available chemotherapeutics to promote sustained release with improved bioavailability [11]. With modified surfaces, these particles interact with proteins in mucus layers and increase their residence time at disease sites [12, 13]. However, the high recycling rate of mucus in the gastrointestinal tract may prevent the loaded nanoparticles from being fully absorbed [14, 15]. Additionally, the adverse environment of the gastrointestinal tract may cause spontaneous dissociation of protein nanocarriers, leading to inadvertent chemotherapeutic release at inaccurate cellular locations, resulting in severe systemic side effects [16]. To maximize drug delivery and minimize unwarranted side effects, drug release using encapsulated enzyme-centred bioorthogonal catalysis (BC) that enables spatiotemporally controlled prodrug activation may increase the efficacy of GC treatment [17, 18]. Catalytic centres, essential for prodrug conversion, are delivered either directly or through vector-mediated gene delivery [19, 20]. However, these applications are limited as harsh gastrointestinal tract conditions may cause their degradation, resulting in low bioavailability and limited absorption of active therapeutics [21, 22].

We introduce our unique oral nanoplatform, disulfide-linked thermostable exoshells (DS-tES), for targeted delivery of BC centres such as horseradish peroxidase (HRP). Structurally modified from the parent molecule (thermostable exoshells or tES) through molecularly precise inter-subunit disulphide linkages via A152C-L53C-G67C-A117C-G37C, DS-tES significantly stabilized the encapsulated catalytic centres in the gastrointestinal tract where tES failed [23]. Like tES, DS-tES is a ferritin-based self-assembling porous exoshell with 24 subunits. It has a hydrodynamic diameter of ~15 nm and an interior aqueous cavity of 8 nm, which can enclose catalytic centres with a theoretical upper limit of ~306 nm³. Additionally, the exoshell contains four surface pores of 4.5 nm, facilitating the passage of small molecules [24, 26]. Unlike tES, DS-tES demonstrates improved intestinal permeability and extended gastrointestinal residence time, as shown in previous studies using Caco-2 cell models and in vivo mouse models, highlighting its potential for oral drug delivery applications [23]. Recently, we described the transport of iron-containing BC centres in HRP to solid tumours via tES, where they catalysed the prodrug IAA, a natural substance present in edible plants, into bioactive metabolites. Encapsulation within tES was crucial for the endocytic uptake of biocatalytic centres, leading to apoptosis and complete regression of an in vivo orthotopic xenograft tumour [17].

Given the current gaps in the oral administration of BC centres, we hypothesize that DS-tES can encapsulate and protect labile reaction centres, such as HRP, from degradation prior to their delivery to GC tissues in vivo, as HRP is known to be unstable under gastrointestinal conditions [23, 27, 28]. In this study, we explored the anti-cancer effects of bioactive metabolites generated through IAA catalysis within DS-tES encapsulated BC centres, thereby targeting this pathway for the prevention and therapy of gastric cancer. Necroptosis is a type of programmed cell death that is linked to a variety of inflammatory disorders [29, 30]. However, cancer cells exhibit minimal expression of necroptotic markers that helps sustained proliferation of cells evading cell death and augments metastasis, indicating the potential of necroptosis inducers as a novel cancer therapeutic approach [31, 35]. In vitro experiments demonstrated that only HRP encapsulated within DS-tES followed by IAA treatment (DS-tES/HRP/IAA) [DTHI) induced necroptosis in both HER2-positive (HER2+) and HER2-negative (HER2-) human GC cells by upregulating necroptotic markers such as receptor-interacting protein 1 (RIP1), receptor-interacting protein 3 (RIP3), and mixed lineage kinase domain-like protein (MLKL). However, in DS-tES (blank nanoparticles), DS-tES/IAA (DS-tES with IAA) [DTI], and DS-tES/HRP (HRP encapsulated within DS-tES with no IAA) [DTH], there was no inhibition of cell proliferation. In a mouse model of gastric cancer induced by N-Methyl-N-Nitrosourea (MNU), oral administration of DTHI, significantly reduced tumour growth without any changes to the normal gastric function. Furthermore, active metabolites from IAA metabolism, such as indole-3-carboxaldehyde (I3A) and indole-3-carbinol (I3C), were quantified in mouse plasma using mass spectrometry and found to abrogate GC cell proliferation in a dose- and time-dependent manner.

Experimental section

Reagents

Major reagents used were as follows: Indole-3-acetic acid (IAA), indole-3-carboxaldehyde (I3A), indole-3-carbinol (I3C), 3-(4,5-Dimethylthiazol-2-yl)−2,5-diphenyltetrazolium bromide (MTT), carbonyl cyanide-p-trifluoromethoxyphenylhydrazone (FCCP), N-Methyl-N-nitroso urea (MNU), β-mercaptoethanol (BME), sodium dodecyl sulfate, bovine serum albumin (BSA), glycerol, dimethyl sulfoxide (DMSO), 1X phosphate-buffered saline with Tween 20 detergent (PBST), phenylmethanesulfonyl fluoride (PMSF), mounting reagent, and cover slips were purchased from Sigma-Aldrich (St. Louis, USA). 100 mm stock concentrations of IAA, I3A, and I3C were prepared in DMSO and diluted to working concentrations in normal saline and pH adjusted to 6.8–7.0. Stable-isotope internal standards such as [D5] Indole-2,4,5,6,7-D5 acetic acid (IAA), [D4] Homocysteine (Hcys), and the amino acid mixture containing [13C,15N] Proline and [13C6,15N2] Cystine were purchased from Scientific Resources Pte. Ltd. (Singapore). Annexin V, nuclease-free water, propidium iodide, Pierce^™ Fast Blocking Buffer, ribonuclease A, trypan blue vital stain, penicillin-streptomycin solution, sodium pyruvate, kanamycin, Tris(2-carboxyethyl) phosphine (TCEP), tetramethylrhodamine ethyl ester (TMRE), 1-Step^™ TMB ELISA substrate solution and TRIzol^™ Reagent were purchased from ThermoFisher Scientific (Waltham, USA). Bradford reagent was purchased from Bio-Rad (Hercules, USA). Hematoxylin, eosin, and clearene were purchased from Leica biosystems (Nussloch, Germany). Blocking One reagent was purchased from Nacalai Tesque (Kyoto, Japan). Absolute ethanol was purchased from Fisher Scientific (Pittsburgh, USA). Isopropyl β-D-1-thiogalactopyranoside (IPTG) and 2X Laemmli buffer was purchased from Axil Scientific (Singapore). Amersham ECL Detection Reagent was purchased from Cytiva Life Sciences (Marlborough, USA). Necrostatin-1 was purchased from Selleck Chemicals (Houston, USA).

Cell lines, NCI-N87 (iCell-h164; RRID: CVCL_1603) and MKN45 (iCell-h345; RRID: CVCL_0434) were purchased from iCell Bioscience Inc. (Shanghai, China). The cell lines were free of contamination. Bacterial Terrific broth (TB) was purchased from Sigma-Aldrich (St. Louis, USA). Cell culture medium RPMI 1640, Phosphate-buffered saline (PBS), fetal bovine serum (FBS), anti-HER2 antibody (Cat# MA5-13105, RRID: AB_10988194), Horseradish peroxidase (HRP) and Goat anti-Rabbit IgG (H + L) Highly Cross-Adsorbed Secondary Antibody, Alexa Fluor^™ 546 (Cat# A-11035, RRID: AB_2534093) were obtained from ThermoFisher Scientific (Waltham, USA). Antibodies for RIPK1 (Cat# 3493, RRID: AB_2305314), phospho-specific RIPK1 (Ser 166) [Cat# 65746, RRID: AB_2799693], RIPK3 (Cat# 13526, RRID: AB_2687467), phospho-specific RIPK3 (Ser 227) [Cat# 93654, RRID: AB_2800206], MLKL [Cat# 14993, RRID: AB_2721822] and phospho-specific MLKL (Ser 358) [Cat# 91689, RRID: AB_2732034], Transferrin receptor/CD71 [Cat# 13113, RRID: AB_2715594], Caspase3 (Cat#14220, RRID: AB_2798429], cleaved caspase 3 [Cat# 9664, RRID: AB_2070042], β-Actin (Cat# 4970, RRID: AB_2223172], GAPDH (Cat# 2118; RRID: AB_561053], anti-rabbit IgG HRP-linked antibody (Cat# 7074, RRID: AB_2099233), anti-mouse IgG HRP-linked antibody (Cat# 7076, RRID: AB_330924) and RIPA buffer were purchased from Cell Signaling Technology (Danvers, USA). Western blot stripping buffer (Cat# AB282569) was purchased from Abcam (Cambridge, UK).

RealTime-Glo^™ Annexin V Apoptosis and Necrosis Assay kit was purchased from Promega Inc, (Madison, USA). MitoPerOx, a fluorescent mitochondria-targeted lipid peroxidation probe and malonyldialdehyde assay kit were purchased from Abcam (Cambridge, UK). BioFlux PCR/RT-PCR kit was purchased from Hangzhou Bioer Technology (Binjiang District, China). PerfeCTa SYBR^® Green FastMix L-ROX, the reaction cocktail for qPCR, was purchased from DiethelmKellerSiberHegner (DKSH) (Zurich, Switzerland).

Preparation of DS-tES

DS-tES shells were produced in an E. Coli expression system as previously described [23]. E. coli competent cells were transformed with a pRSF plasmid positive for the DS-tES gene, and glycerol stocks of these cells were made. An overnight starter culture (25 ml), prepared from the glycerol stock, was used to inoculate TB media (1 L) supplemented with kanamycin (50 mg/ml). The culture was maintained at 37 °C until the absorbance (OD600) reached between 0.5 and 0.6. Protein expression was then induced by adding IPTG (0.5 mm), and the culture was allowed to grow for another 24 h at 37 °C. The bacterial cells were pelleted, resuspended in lysis buffer (50 mm Tris-HCl, pH 8.0, 150 mm NaCl, 5 mm BME, and 0.1% Triton-X 100) and sonicated to obtain soluble protein fractions. The soluble fractions were heated at 60 °C for 5 min to precipitate out all other bacterial proteins, after which they were spun down and filtered. To obtain purified DS-tES shells, these filtered fractions were subjected to size-exclusion chromatography (SEC) column equilibrated with 50 mm Tris, pH 8.0 and BME (5 mm). The purity of the proteins was evaluated using SDS-PAGE, and the proteins were stored under reducing conditions for long-term storage. For cellular uptake studies, DS-tES conjugated with Alexa 488 (Alexa 488 C5 maleimide; Thermofisher Scientific, Waltham, USA) was prepared according to the manufacturer’s instructions.

Characterization of DS-tES

The morphologies of DS-tES at pH 4.0 and pH 7.4 were examined via transmission electron microscopy (TEM) (Thermofisher Talos L120C) at the Department of Medicine, National University of Singapore, TEM facility. Particle size and distribution of DS-tES at both pH values were assessed in triplicates by dynamic light scattering (Zetasizer Lab, Malvern Panalytical), using disposable cuvettes at 24 °C. Zeta potential values at the respective pH levels were also determined using the Zetasizer Lab (Malvern Panalytical).

Preparation of DTH

DTH was prepared as previously described [23]. The reducing agent TCEP (30 mm) was added to DS-tES, and the solution was acidified to pH 5.8 to dissociate the shell into its subunits. The subunits were mixed with HRP powder in a 20-fold mole excess relative to DS-tES. The mixture was incubated for 10 min at room temperature, after which the pH was adjusted back to 8.0. This mixture was purified with SEC, with the column equilibrated with 1X PBS. The protein concentration was quantified via Bradford assay, while the activity of encapsulated HRP was measured using a 1-Step^™ TMB ELISA assay. The morphologies and particle size distribution of DTH at pH 4.0 and pH 7.4 were analyzed in the same manner as DS-tES.

Determining kinetic parameters of free HRP and encapsulated HRP

The enzyme kinetics of encapsulated HRP were compared to free HRP by using a fixed concentration of hydrogen peroxide and varying concentrations of TMB substrate (10 µm to 400 µm). The concentrations of free HRP and encapsulated HRP used were fixed at 0.5 nm. Absorption was measured at 652 nm continuously for 15 min. Initial reaction rates (V₀) of the TMB substrate were determined in µm per second from the absorbance-time curves using the Beer-Lambert law (ε₆₅₂ = 39000 M^–1 cm^–1) [36]. The kinetic parameters K_M (Michaelis constant), k_cat (turnover number), and catalytic efficiency (k_cat/K_M) for both forms of HRP were determined to provide a detailed comparison.

Cell culture

NCI-N87 and MKN45 cell lines were cultured at 37 °C in a 5% CO₂, 95% O₂ atmosphere in a Thermo Forma incubator (Thermo Fisher). The cells were maintained in RPMI-1640 medium supplemented with fetal bovine serum (10%), sodium pyruvate (1%), and penicillin-streptomycin (1%). The culture medium was changed every two days until 80–90% confluency was reached.

Cell viability assay

The antiproliferative effects of DTH in combination with IAA were determined using an MTT assay as described previously [37]. Cells (1 × 10⁴ cells/well) were seeded into a 96-well plate and grown at 37 °C with a 5% CO₂ atmosphere. DTH in combination with IAA was added to each well such that the final concentration of IAA was 1 mm and encapsulated HRP was 2.5 µg. Similarly, the antiproliferative effects of I3A and I3C were investigated using the same method, at doses ranging from 0.75 mm to 0.046 mm. After treatment, the plate was incubated for 24, 48 and 72 h. At each designated time point, 20 µl of MTT solution (5 mg/mL) was added to each well, and the plate was incubated for 4 h. After the incubation, the media in each well was replaced with 100% DMSO (100 µl) to dissolve the formazan crystals. After 10 min, the absorption was measured at the wavelength of 570 nm in a 96-well plate reader (Tecan, USA).

Analysis of cellular DNA content by flow cytometry

Cells (1 × 10⁵ cells/well) were seeded on 6-well cell culture plate, with and without DTHI (2.5 µg encapsulated HRP and 1 mm IAA) for the indicated time points. After incubation, the cells were collected, washed thrice with 1X PBS and then fixed in ice cold ethanol (70%) for 1 h on ice. Cells were pelleted at 1000 g for 5 min and resuspended on ice in 1X PBS containing PI (5 µg/ml) and ribonuclease A (1 mg/ml) for 10 min. The cellular DNA content of the cells was analyzed using CytoFLEX flow cytometer software (Beckman Coulter, USA). Hypodiploid cells, due to DNA fragmentation, accumulate in the sub-G1 phase of the cell cycle and were thus regarded as apoptotic cells.

Annexin V/PI double staining

Apoptosis and necrosis were differentiated using Annexin V/PI double staining, where annexin V positive/PI negative cells indicate apoptosis, while PI positive cells indicate necrosis [38]. Cells (1 × 10⁵ cells/well) were plated in a 6-well plate and cultured overnight. The cells were then treated with DTHI (2.5 µg encapsulated HRP and 1 mm IAA) for the indicated time points. After treatment, the cells were trypsinized, washed three times with ice-cold 1X PBS and then incubated in annexin V-FITC/PI binding buffer containing Annexin V staining dye (5 µL) for apoptosis detection. PI (5 µg/ml) was also added to the cell suspension, and the mixture was incubated at room temperature in the dark for 15 min. The cells were then filtered and analyzed using the CytoFLEX flow cytometer software (Beckman Coulter, USA). Fluorescence was detected at wavelengths of 488 nm and 617 nm.

Apoptosis/Necrosis assay

Cells were seeded in a sterile, transparent bottom white 96-well tissue culture plate at a concentration of 10,000 cells/100 µL volume/well. Both the GC cell lines were incubated at 37 °C with 5% CO₂. After 24 h, cells were treated with or without DTHI (2.5 µg encapsulated HRP and 1 mm IAA) for 6, 12, 24, and 48 h. The cells were also treated with or without DTHI at the same concentration in the presence of Nec-1. Analysis was performed with a RealTime-Glo Annexin V Apoptosis and Necrosis Assay kit according to the manufacturer’s protocol. This assay measures the exposure of phosphatidylserine, a phospholipid found on the outer leaflet of the cell membrane, during early apoptosis and absent during secondary necrosis, which is characteristic of the loss of cell membrane integrity. Annexin V bound to phosphatidylserine was detected as a luminescence signal, while necrosis was detected by a profluorescent DNA dye. The luminescence and fluorescence signal were measured on a 96-well plate reader (Tecan, USA).

Mitochondrial membrane potential assay

At the designated treatment time, TMRE dye (final concentration of 25 nm) was added to both the DTHI treated cells (2.5 µg encapsulated HRP and 1 mm IAA) and control cells. The cells were incubated for 30 min at 37 °C in a 5% CO₂, 95% O₂ atmosphere. A positive control FCCP (final concentration of 20 µm), an uncoupler of mitochondrial oxidative phosphorylation, was also added for 30 min. The cells were then washed with 1X PBS, trypsinized, pelleted, and resuspended in 1X PBS (200 µL). A CytoFLEX flow cytometer (Beckman Coulter, USA) was used to analyze the cells set at an excitation/emission wavelength (Ex:549/Em:575 nM).

Lipid peroxidation assay

Lipid peroxidation was assessed using the MitoPerOx dye, a mitochondrial-targeted derivative of the C11-BODIPY (581/591 nM) probe, selected for its efficient mitochondrial uptake. Upon lipid peroxidation, the dye exhibits a fluorescence emission shift from ~590 nM to ~520 nM when excited at 495 nM. Following the designated treatment period, MitoPerOx (final concentration of 2 µm) was added to both DTHI-treated cells (containing 2.5 µg encapsulated HRP and 1 mM IAA) and control cells. After a 15-min incubation, cells were washed with 1X PBS, trypsinized, and centrifuged. The resulting cell pellets were resuspended in 1X PBS (200 µl) and analyzed using a CytoFLEX flow cytometer (Beckman Coulter, USA).

Malondialdehyde assay

Cells (1 × 10⁶ cells/well) were plated in a 6-well plate and treated with DTHI (2.5 µg encapsulated HRP with 1 mM of IAA) for up to 48 h. After treatment with DTHI, the cells were washed with 1X ice cold PBS and lysed using MDA lysis buffer. The cells were homogenized, centrifuge at 13,000 g for 10 min and the supernatant containing the cell lysate was collected for the assay. 20 µL of each sample was mixed with H₂SO₄ (42 mm), followed by the addition of phosphotungstic acid solution (125 µL) to precipitate lipids. This mixture was incubated for 5 min at room temperature and then centrifuged at 13,000 g for 5 min. The cell pellet was resuspended on ice with deionized water (100 µL) containing butylated hydroxy toluene (2 µL). 200 µL of each cell suspension was transferred into a 96-well plate. The concentration of MDA was then estimated via measuring absorbance at a wavelength of 532 nm using a 96-well plate reader (Tecan, USA).

Western blotting

Lysates from MKN45 cells or gastric tissues across treatment groups were prepared in RIPA buffer supplemented with 1 mM PMSF, following the manufacturer’s protocol. Protein concentrations were quantified using the Bradford assay, and 50 µg of total protein per sample was subjected to Western blot analysis. Samples were mixed with 2× Laemmli buffer, heated at 95 °C for 5 min, and resolved on a 12% SDS-PAGE gel at 125 V for 60 min. Proteins were transferred to a nitrocellulose membrane using the Invitrogen^™ iBlot 3 Western Blot Transfer System. Membranes were blocked with Pierce^™ Fast Blocking Buffer for 1 h at room temperature and incubated overnight at 4 °C with primary antibodies against RIPK1, RIPK3, MLKL, β-actin (1:1000) and GAPDH (1:1000), phospho-RIPK1 (Ser166), phospho-RIPK3 (Ser227), phospho-MLKL (Ser358) (1:500), Caspase-3 (1:500), and cleaved Caspase-3 (1:500). Following two washes with PBST, membranes were incubated with HRP-conjugated anti-mouse or anti-rabbit secondary antibodies (1:5000) for 1 h at room temperature, washed again, and developed using Amersham^™ ECL detection reagents on a Bio-Rad ChemiDoc Imaging System. Membranes were subsequently stripped and reprobed for β-actin or GAPDH as loading controls.

qPCR

Both MKN45 and NCI-N87 cells were lyzed using TRIzol^™ Reagent, following the manufacturer’s instructions. RNA was extracted and concentration was quantified and normalized to 1 µg by diluting with Nuclease-Free water. cDNA was synthesized using commercially available cDNA synthesis kit according to manufacturer instructions [BioFlux PCR/RT-PCR kit (China)]. Primers were dissolved in Nuclease-Free water to achieve a final stock concentration of 100 mm (Supplementary Table 1 for primer details). qPCR was then performed using a PerfeCTa SYBR Green FastMix, L-ROX qPCR kit, according to the manufacturer’s instructions.

Cell uptake studies

Transferrin receptor (TFR) expression levels in gastric cancer cells were assessed by western blot using a TFR antibody (1:500) as outlined in Sect. "Western blotting". MKN-45 gastric cancer cells were seeded onto 8-well chamber slides (ibidi, Cat. No. 80826) and cultured overnight to allow for cell attachment. The following day, live cells were treated with either unconjugated DS-tES or DS-tES conjugated to Alexa Fluor 488 for 2 h. For receptor-blocking studies, cells were pre-treated with TFR antibody (1:500) overnight at 4 °C prior to incubation with Alexa Fluor 488-conjugated DS-tES. After treatment, cells were washed with PBS and blocked with 3% BSA in PBS for 1 h at room temperature. They were then incubated with Goat anti-Rabbit IgG secondary Antibody (1:1000), Alexa Fluor^™ 546 for 1 h at RT, followed by PBS washes. Nuclei were counterstained with Hoechst (1 µg/mL) for 30 min at RT, washed twice with PBS and mounted. Fluorecence imaging was performed using an Olympus FV3000 microscope.

Animal study

Mice (n = 30) were given 240 ppm (parts per million) of MNU in drinking water in light shielded bottles for 10 weeks (benign hyperplasia) followed by a higher dose of 480 ppm for another 10 weeks (development of polyps and adenoma). Following MNU administration, the mice were randomly assigned to 3 treatment groups (n = 10/group). Group 1: Vehicle control (50 mg/kg DS-tES), Group 2: DTI (blank DS- tES administered by oral gavage at a dose of 50 mg/kg followed by 400 mg/kg IAA), and Group 3: DTHI (DS-tES/HRP administered by oral gavage at a dose of 50 mg/kg total protein containing 150 µg of HRP followed by 400 mg/kg IAA). The dose of IAA was chosen based on toxicity and tolerability studies [39]. Treatment began on the 21 st week after MNU administration. Body weight of the mice was measured weekly throughout the study. At the end of the study, the mice were sacrificed, and their stomachs were harvested to assess and quantify various tumour phenotypes such as hyperplasia, hyperplasia with polyps and adenocarcinoma. The number of tumours in the stomach was counted using a stereomicroscope.

Histology studies

Haematoxylin-Eosin staining was conducted as previously described [40, 41]. Briefly, stomach tissue from control and DTHI-treated mice was fixed in 10% phosphate buffered formalin and embedded in paraffin blocks. 5 μm sections were deparaffinized in Clearene (xylene substitute), dehydrated in graded alcohol and rehydrated in water. The sections were stained with Harris haematoxylin for 5 min, rinsed in water and then stained with 1% eosin for 3 min. After staining, the sections were dehydrated using graded ethanol and Clearene. The mounted slides were then examined, and images were captured using Leica THUNDER microscope (magnification, 10X and 20X).

Indole-3-acetic acid, Indole-3-acetaldehyde and Indole-3-carbinol quantification

A linear calibration curve was constructed for IAA by plotting analyte-to-internal standard peak area ratios (As/Ai) against molar concentrations of IAA, with a serial dilution factor of 2. At each concentration level, measurements were taken at least three times, with the highest and lowest concentrations sampled five times each. The calibration curves for I3A and I3C were constructed using the same method as for IAA.

Panel-based LC-MS/MS metabolomic assay

The LC-MS/MS assay was performed with the Agilent 1290 Infinity II/Agilent 6495 C Triple Quadrupole system with both positive and negative electrospray ionization (ESI). Plasma samples of 50 µL were transferred into 96-well plates and were spiked with 50 µL of internal standard mixtures (5 µg/L) to account for assay variability and control for signal-concentration linearity (Supplementary Text 1). IAA metabolites and other annotated metabolites were monitored [42, 43]. The LC was performed using peek-coated SeQuant^®ZIC^®-cHILIC 3 μm, 100 Å 100 × 2.1 mm HPLC column (Merck Pte Ltd., Singapore) with zwitterionic phosphorylcholine head groups maintained at 40 °C. The organic solvent was acetonitrile containing 0.1% formic acid (solvent A) and the aqueous solvent used was 20 mm ammonium formate pH 4.0 (solvent B). A linear LC gradient on binary pump (Agilent Model G7120A) was set up with the percentage of solvent B as follows: 10% at 0 min, 70% at 9.00 min, 70% at 11.0 min, and 10% between 11.1 and 11.5 min, with a flow rate of 0.4 mL/minute. The column was further equilibrated for another 11.5 min with 10% solvent B. An additional high-speed pump, binary pump B (Agilent Model G7120A), together with a quick-change valve head, 2-position/10-port, 1,300 bar (Agilent Part No: 5067 − 4240), were utilized. The percentage of solvent B on binary pump B was maintained at 10% with a flow rate of 0.4 mL/minute. The sample injection volume was 10 µL. The MS conditions were as follows: capillary voltage of 4.0 kV, nozzle voltage of 500 V, iFunnel parameter high/low-pressure RF of 90 V, nebulizer pressure of 60 psi, the gas temperature of 290 °C, sheath gas temperature of 350 °C, nebulizer was 35 psi and sheath gas flow of 12 L/minute.

Software and statistical analysis for the LC-MS/MS metabolomic assay

Chromatograms were acquired with the Agilent dynamic multiple-reaction monitoring (dMRM) scan mode and was implemented with the Agilent MassHunter Workstation Acquisition (10.0.127) software. Areas under the peaks were quantified with the Agilent MassHunter Quantitative Analysis (10.1.733) software. Statistical analysis and data preprocessing were performed with R (4.3.1) and MetaboAnalyst (v6.0) [44]. To account for sensitivity drift, the relative abundance of metabolites was quantified using the response ratio approach, which normalizes the peak area of each analyte by the peak area of the corresponding deuterated internal standard. All samples were analyzed in random order. To account for variances in sample dilution, sample-wise probabilistic quotient normalization was performed using the median concentration measured for each sample and across all metabolites. High-variant metabolites with the highest 5% of interquartile ranges were filtered. Metabolite readings were further mean-centred and standardized before group-wise comparisons were performed.

Statistical analysis

All numerical data were expressed as means±SD. Unpaired t-tests or Mann-Whitney U tests were performed for group-wise comparison where appropriate. P < 0.05 was considered statistically significant.

Results

Encapsulation of bioorthogonal catalytic centres have a minimal impact on its catalytic activity

We previously established that tES positive (tES (+)) shells can encapsulate and stabilize HRP molecules at single or double stoichiometric ratios through charge complementation between HRP’s negative surface charge and the net positive charge of the tES(+) internal cavity [17, 25]. DS-tES, derived from tES (+) for enhanced stability across a wide pH range, was assessed for its morphology, stability, and surface potential under both gastric and intestinal pH conditions. Considering the average gastric pH of 4.0 in the fed state [45], the stability of DS-tES was assessed using transmission electron microscopy (TEM) and dynamic light scattering (DLS). Similar to assembled tES, DS-tES exhibited a spherical morphology at pH 4.0. The TEM images revealed a hollow nanoshell structure, with an internal diameter of 8 nm and an approximate outer diameter of 13 nm, with comparable morphology also observed at pH 7.4. However, the net surface charge of DS-tES shifted from negative to positive when the pH was lowered to 4.0, consistent with previous observations [46]. At pH 4.0, DS-tES displayed a positive zeta potential of +14.59 ± 0.5 mV, a shift from the negative zeta potential of −27.52 ± 1.4 mV observed at pH 7.4. Similarly, at pH 4.0, DS-tES had a hydrodynamic diameter of 20.9 ± 1.4 nm with an average polydispersity index of 0.2, whereas at pH 7.4, the hydrodynamic diameter was 15.8 ± 0.5 nm with an average polydispersity index of 0.3 (Fig. 1). Recent studies investigating pH dependent structural changes in ferritin have reported similar size values (~20 nm) at acidic pH, likely due to conformational changes [47, 48]. Importantly, TEM and DLS analyses revealed no significant changes in the morphology or hydrodynamic diameter of DTH, which remained consistent with HRP encapsulated within tES, indicating that DS-tES’s structural integrity was maintained despite the mutations at the subunit interface (Supplementary Fig. 1) [23, 26].

Fig. 1.

(A) and (B) (enlarged view). Structural representation of DS-tES (blue) with 4.5 nm triangular surface pore showing the encapsulated bioorthogonal reaction centre. The reaction centre consists of a heme prosthetic group as the transition metal complex (cyan) protected by the coordinating enzyme HRP (magenta). Yellow represents the cysteine residues on the exoshell. PyMOL was used for the structural visualization of the encapsulated DS-tES. (C) and (D) TEM analysis of DS-tES at pH 4.0 and 7.4, respectively showing spherical morphology of the assembled nanoparticles. (E) and (F) Size-distribution of DS-tES at pH 4.0 and 7.4 showing hydrodynamic diameter of 20.9 nm and 15.8 nm, respectively. (G) Zeta potential of DS-tES at pH 4.0 and 7.4. At pH 4.0, DS-tES exhibited mean surface potential of +14.59 mV and at pH 7.4, DS-tES exhibited mean surface potential of −27.52 mV

Furthermore, the catalytic activity of DTH on 3,3′,5,5′-tetramethylbenzidine was evaluated and compared to that of free HRP. Importantly, DTH was significantly active and maintained the ability to catalyse small molecular substrates, potentially via the four 4.5 nm surface pores in the shell (Supplementary Fig. 2, Supplementary Table 2) [25]. The kinetic parameters k_cat (turnover number) and k_cat/K_M (catalytic efficiency) for DTH and free HRP were determined as 460.6 ± 14.9 S^{− 1} and 792.4 ± 41.8 S^{− 1}; 3.4 ± 0.2 µm^{− 1}S^{− 1} and 7.1 ± 0.4 µm^{− 1}S^{− 1}, respectively. While enzyme kinetics data indicate that the porous nature of DS-tES allows substrate access, the slight difference in the HRP’s catalytic ability post-encapsulation could be possibly due to diffusional limitations of the small molecular substrates imposed by the surrounding cage [49, 50].

DTHI inhibits the proliferation of gastric cancer cell lines

As HER2- GC represents majority of the cases (80–88%) [51, 52], we investigated the inhibitory effects of DTHI on cell viability in both HER2- (MKN45) and HER2+ (NCI-N87) human GC cell lines. DTHI significantly inhibited cell growth in a dose- (Fig. 2A and B) and time-dependent manner (Fig. 2C and D). Exposure to DTHI (2.5 µg of HRP encapsulated within DS-tES followed by 1 mm IAA treatment) for 48 h inhibited MKN45 and NCI-N87 cell proliferation by 65% and 68%, respectively (Fig. 2A and B). Similar observations were also made by Samira Sadeghi et al., when HRP encapsulated within tES(+) shells efficiently catalysed IAA, resulting in a 70% reduction in cell viability of MDA-MB-231 triple negative breast cancer cells [17]. Necrostatin (Nec-1, 100 µm), a RIP1-specific inhibitor of necroptosis, reduced DTHI-induced cell death and increased cell proliferation by 62% in MKN45 cells and 69% in NCI-N87 cells, suggesting that the necroptosis pathway was involved in DTHI-induced cell death. Furthermore, no cell death was observed when treated with either DTI or DTH alone (Fig. 2C and D), demonstrating that the combination of HRP and IAA generated cytotoxic metabolites that inhibited cancer cell proliferation regardless of HER2 expression. IAA was previously reported to be an inert compound [53, 54]. In our study, treatment with IAA alone did not exhibit any cytotoxic effects (data not shown). Western blot image showing the presence of HER2 protein in NCI-N87 cells with 1.42 fold increase in the protein levels compared to MKN45 cells which does not express HER2 protein (Supplementary Fig. 3).

Fig. 2.

Cytotoxic effects of DTHI (DS-tES/HRP/IAA) on gastric cancer cell lines as determined by MTT assay. (A) and (B) MKN45 cells (left panel) and NCI-N87 cells (right panel) were treated with various concentrations of DTH (HRP encapsulated within DS-tES) and co-incubated with a single dose of IAA (1 mm) for 24 h and 48 h. (C) and (D) Time course study where MKN45 cells (left panel) and NCI-N87 cells (right panel) were treated with single dose of DTHI (DS-tES with 2.5 µg of encapsulated HRP and 1 mm IAA) either in the presence or absence of 100 µm Nec-1. DS-tES (blank nanoparticles), DTI (DS-tES with 1 mm IAA and without HRP), and DTH (DS-tES with 2.5 µg of encapsulated HRP and without IAA) served as controls. Data shows mean±SD (n = 6). *P < 0.05; **P < 0.005; #P < 0.05; ##P < 0.005

DTHI induce necroptosis in gastric cancer cell lines

DTHI treatment resulted in a significant accumulation of apoptotic cells in the sub-G1 phase of the cell cycle in both cell lines, as demonstrated by increased DNA fragmentation (Fig. 3A and B, Supplementary Fig. 4). When necroptosis was assessed using annexin V/propidium iodide (PI) staining, 44% of MKN45 (Fig. 3C and D) and 46% of NCI-N87 cells (Fig. 3E and F) were PI positive after 48 h of treatment. However, Nec-1 administration reduced cells undergoing necroptosis in MKN45 and NCI-N87 cells by 23% and 17%, respectively, indicating that DTHI is a strong inducer of necroptosis in gastric cancer cells. Notably, Nec-1 had no effect on early apoptosis and was not statistically significant.

Fig. 3.

(A) and (B) Time-dependent accumulation of apoptotic cells in the sub-G1 phase of the cell cycle in both MKN45 (left panel) and NCI-N87 (right panel) cells treated with DTHI (DS-tES with 2.5 µg of encapsulated HRP and 1 mm IAA). (C) and (D) MKN45 and (E) and (F) NCI-N87 cells were treated with DTHI for 48 h. After 48 h, the cells were stained with Annexin V and PI. FACS analysis was performed to determine the proportion of early apoptotic and necroptotic cells. C. and E. Representative flow cytometry histogram of MKN45 and NCI-N87 cells, respectively. D. and F. Percentage cell population of MKN45 and NCI-N87 cells, respectively. Each group was analyzed in triplicate. *P < 0.05; **P < 0.005; # P < 0.05

To investigate whether necroptosis was involved in DTHI induced cell death, we used RealTime Glo^™ Annexin V Apoptosis and Necrosis assay kit. A comprehensive examination of programmed cell death using a combination of luminescent (Annexin V binding) and fluorescent (DNA release) responses in the presence of DTHI revealed plasma membrane permeabilization, leading to the leakage of intracellular contents and triggering necroptotic cell death. While DTHI exhibited early apoptosis through a time-dependent increase in luminescence from phosphatidylserine (PS) exposure in both cell lines, a drop-in response at 48 h indicated a loss of cell membrane integrity. Similarly, necrosis observed by an increase in fluorescence between 24 and 48 h, was detected using a profluorescent DNA binding dye that specifically detects necrotic cell death. Collectively, DTHI triggered necroptosis in both HER2- and HER2 + expressing gastric cancer cell lines. Nec-1 suppressed DTHI-induced necroptosis, as indicated by increase in luminescence and decrease in fluorescence responses (Fig. 4A and B).

Fig. 4.

RealTime Glo^™ Annexin V Apoptosis and Necrosis assay was performed as per manufacturer instructions as described in materials and methods. (A) MKN45 cells and (B) NCI-N87 cells were exposed to a single dose of DTHI (DS-tES with 2.5 µg of encapsulated HRP and 1 mm IAA) for various time points. Relative luminescence units (phosphatidylserine: annexin V binding) and relative fluorescence units (membrane integrity) were collected at indicated times. Data represents the mean±SD (n = 6), #P < 0.05; ^{# #}P < 0.005

DTHI induced RIP1-dependent necroptotic pathway

The key mechanism of necroptosis is related to the activation and phosphorylation of receptor interacting protein kinase-1 and −3 and mixed lineage kinase domain-like protein (RIP1/RIP3/MLKL) pathway. We next investigated the effect of DTHI on the expression and phosphorylation of these necroptotic factors in both the GC cell lines. RIP1 phosphorylation recruits RIP3, resulting in the formation of RIP1-RIP3 complex or necrosome, which then phosphorylates and oligomerizes MLKL. The oligomerized MLKL translocate to the cell membrane and subsequent permeabilization results in necroptosis [55, 56].

DTHI significantly upregulated the RIP1-dependent pathway in both the cell lines evaluated [56]. qPCR analyses post DTHI treatment in both the GC cell lines revealed time-dependent upregulation of RIP1 (Fig. 5A and D), RIP3 (Fig. 5B and E) and MLKL (Fig. 5C and F) mRNA transcripts at 48 h, thereby indicating the induction of necroptosis pathway mediated cell death. Interestingly, Nec-1 downregulated RIP1 and RIP3 and did not have any effect on the upregulation of MLKL (data not shown). Nec-1 is known to negatively regulate RIP1, inhibiting its dimerization with RIP3, and subsequently downregulating RIP1/RIP3/MLKL signalling pathway [55, 57, 59]. Furthermore, Western blot analysis with MKN45 cells indicated that DTHI upregulated RIP1, RIP3, and MLKL, as well as its phosphorylation (Fig. 5G). We observed that Nec-1 decreased the expression and phosphorylation of RIP kinases while having no effect on MLKL expression (Fig. 5G). Overall, our results demonstrate that DTHI treatment promoted necroptotic cell death in gastric cancer cells.

Fig. 5.

qPCR analysis. Gastric cancer cells MKN45 (top panel) and NCI-N87 (bottom panel), were treated with DTHI (DS-tES with 2.5 µg of encapsulated HRP and 1 mm IAA) for the indicated times either in presence or absence of Nec-1. qPCR analysis was performed and the mRNA expression levels of RIP1 (A. and D.) RIP3, (B. and E.), and MLKL (C. and F.). G. Western blot analysis was performed as described in materials and methods. DTHI induces RIP1, RIP3 and MLKL expression and phosphorylation in gastric cancer cells in a time-dependent manner. Results showed an increase in RIP1/phospho-RIP1(Ser166), RIP3/phospho-RIP3(Ser227), and MLKL/phospho-MLKL(Ser358) expression in MKN45 gastric cancer cells. Treatment with Nec-1 reduced the expression of key necroptotic proteins, suggesting the involvement of necroptosis. Data represents the mean±SD (n = 6), *P < 0.05; **P < 0.005; ^#P < 0.05; ^{# #}P < 0.005

DTHI induce lipid peroxidation in gastric cancer cells

The reactive intermediates produced by oxidative decarboxylation of IAA have been shown to stimulate ROS production, which has been linked to membrane lipid peroxidation and cellular damage [53, 60, 61]. As a result, we investigated the effect of DTHI treatment on lipid peroxidation in mitochondria using a radiometric fluorescent mitochondria-targeted lipid peroxidation probe. Quantitative examination revealed a time-dependent increase in fluorescence response, indicating higher lipid peroxidation in the presence of DTHI compared to untreated cells (Fig. 6A and B). Malondialdehyde (MDA) is the final product of lipid peroxidation, and its abundance correlates with the extent of oxidative cellular damage [62, 63]. We next investigated the effect of DTHI on MDA production in gastric cancer cells. DTHI induced time-dependent increase in MDA levels in cellular lysates compared to untreated cells (Fig. 6C and D). We believe that an increase in MDA levels can cause endoplasmic reticulum stress and cell death, implying that ER stress initiates the unfolded protein response, which then leads to necroptosis in GC cell lines [64, 65].

Fig. 6.

Mitochondrial lipid peroxidation assay. A. and B. MKN45 and NCI-N87 cells treated with a single dose of DTHI (DS-tES with 2.5 µg of encapsulated HRP and 1 mm IAA) for different time points and stained with MitoPerOx dye, respectively. Upon oxidation, the fluorescence emission at 520 nm increases and that at 590 nm decreases, making the 520/590 fluorescence intensity ratio a marker of mitochondrial lipid peroxidation within cells. C. and D. MDA levels in cell lysates of MKN45, and NCI-N87, respectively. Data represents the mean±SD (n = 3–8), *P < 0.05; **P < 0.005

DTHI induce mitochondrial depolarization in gastric cancer cells

As necroptosis is often associated with loss of MMP and increased accumulation of reactive oxygen species (ROS), we evaluated the mitochondrial activity in DTHI-treated gastric cancer cells by TMRE staining, in which the intensity of the fluorescence was proportional to membrane potential. As shown in Fig. 7, DTHI significantly reduced TMRE fluorescence in a time-dependent manner compared to the untreated group, indicative of mitochondrial depolarization. In addition, MDA from lipid peroxidation may also deactivate mitochondrial enzymes, enhance intracellular ROS generation, and decrease membrane potential, eventually resulting in mitochondrial dysfunction and cell death [66, 68].

Fig. 7.

Mitochondrial membrane potential assay. TMRE fluorescence was quantified by flow cytometry. Gastric cancer cells were treated with a single dose of DTHI (DS-tES with 2.5 µg of encapsulated HRP and 1 mm IAA) for different time points and stained with TMRE as described in the materials and methods. (A) Representative flow cytometry histogram of MKN45 cells treated with DTHI at different time points. FCCP is used as the positive control. (B) Geometric mean fluorescent intensity of MKN45 cells. (C) Representative flow cytometry histogram of NCI-N87 cells treated with DTHI at different time points. (D) Geometric mean fluorescent intensity of NCI-N87 cells. Data represents the mean±SD (n = 3), *P < 0.05; **P < 0.005

Transferrin receptors mediate cellular uptake and internalization of DS-tES

To evaluate whether DS-tES directly interacts with host-cell TFR, cells were either co-incubated with fluorescently labeled DS-tES-488 or pre-treated with a TFR-specific antibody prior to DS-tES-488 exposure. Confocal microscopy demonstrated that blocking TFR markedly reduced DS-tES binding to the plasma membrane and prevented its internalization, indicating that TFR serves as the primary receptor facilitating DS-tES transport across the membrane (Fig. 8C and D). In contrast, in the absence of TFR blockade, DS-tES-488 exhibited robust membrane binding, internalization, and cytoplasmic localization, further supporting TFR-dependent cellular entry (Fig. 8A and B).

Fig. 8.

Confocal imaging analysis of receptor-mediated DS-tES uptake under various treatment conditions. A. Cells exposed to non-fluorescent DS-tES exhibited no detectable signal. B. Exposure to DS-tES-488 resulted in distinct plasma membrane localization followed by intracellular accumulation, indicative of active uptake. C. Cells treated with fluorescently labelled TFR antibody displayed strong fluorescence corresponding to receptor overexpression. D. Pre-blocking TFR with the antibody prior to DS-tES-488 addition markedly reduced or eliminated DS-tES internalization, demonstrating that DS-tES uptake is largely dependent on unblocked TFR. Scale bars, 20 μm

DTHI inhibited N-Methyl-N-Nitrosourea induced gastric cancer in vivo

MNU is a potent alkylating agent that induces somatic mutations in gastric epithelial cells and drives the stepwise development of gastric cancer, progressing through intestinal epithelial hyperplasia, heterogeneous hyperplasia, and chronic atrophic gastritis, hallmarks of gastric precancerous lesions [69, 70]. Consistent with these pathological changes, MNU-treated mice displayed mildly rough gastric surfaces with scattered haemorrhages, swollen mucosa, abnormal glandular architecture, and reduced tissue flexibility.

When administered ad libitum in drinking water, MNU reliably produced high-incidence gastric malignancies in the antrum, accompanied by granular alterations, erosion, submucosal vascular hyperplasia, and polyp formation [69, 71]. In this model, microscopic examination revealed that DS-tES–treated mice exhibited inflammation-associated redness, angiogenesis, dysplasia, and small polyps in the antrum, whereas DTHI treatment markedly attenuated these features (Fig. 9B). Because IAA alone lacks cytotoxicity and free HRP is unstable under gastrointestinal conditions [23, 72], groups receiving HRP + IAA or IAA alone were omitted. Histopathological analyses (H&E staining) showed that DS-tES–treated mice developed 6–9 antral polyps on average, while DTHI substantially reduced tumour burden by suppressing MNU-induced polyp growth (Fig. 9C and D), with no detectable injury to small or large intestinal tissues (data not shown). These in vivo findings also confirmed that DS-tES effectively stabilized encapsulated HRP under gastric conditions, in agreement with earlier data demonstrating that DTH retained approximately 14-fold higher enzymatic activity than free HRP under simulated gastric acidity with pepsin [23]. Free HRP remained highly susceptible to proteolytic degradation, as supported by Supplementary Fig. 5. Mechanistic analyses further revealed that DTHI did not activate cleaved caspase-3, excluding caspase-dependent apoptosis; instead, DTHI markedly increased RIP1/p-RIP1 and RIP3/p-RIP3 expression, indicating necroptosis as the dominant cell-death pathway (Fig. 9E and F). Throughout the study, MNU exposure did not significantly affect body weight, and treatments with DS-tES, DTI, or DTHI caused no substantial weight loss, demonstrating good tolerability of the oral nanoparticle formulations in the animals without overt toxicity (Fig. 9G).

Fig. 9.

In vivo assay. A. Schematic of the experiment design. Mice were administered 240 ppm MNU in drinking water for 10 weeks, followed by 480 ppm of MNU in drinking water for another 10 weeks. Treatment with DTHI (DS-tES [50 mg/kg] with encapsulated HRP [150 µg] followed by IAA [400 mg/kg] b.w) was started on week 21 and continued for 6 weeks. B. Representative images of (i) MNU-treated whole stomach showing redness in the antrum, (ii) DS-tES treated stomach section showing dysplasia, polyp, and redness (iii) DTI-treated mice stomach, (iv) DTHI-treated mice stomach, and (v) normal mice stomach. C. Charts showing the number of polyps in the antrum stomach in the control and DTHI-treated groups. D. H&E stained sections of the antrum stomach of control and DTHI-treated mice. Representative images of (i) normal stomach, (ii) DS-tES treated, (iii) DTI treated, and (iv) DTHI treated. M – mucosa; SM – sub mucosa; MP – muscularis propria; MM – muscularis mucosa; P – polyp. E. Western blot analysis of gastric tissues revealed that DTHI treatment upregulated necroptotic markers, including increased expression and phosphorylation of RIP1 and RIP3 F. Pro-caspsase 3 was detected and was not activated as evidenced by the absence of cleaved caspase 3. NT: normal gastric tissue; DT: DS-tES treated; DTI; DTHI treatment groups. G. Mice body weight determined on weeks 20 and 26 showed no significant changes. Data represents the mean±SD (n = 6), **P < 0.005; ^{# #}P < 0.005; n.s - nonsignificant. Scale bars, 50 μm

Detection of IAA metabolites in mice plasma

Utilizing liquid chromatography coupled to mass spectrometry (LC-MS/MS), we determined the plasma levels of IAA and its active metabolites I3A and I3C, formed through oxidative decarboxylation [53, 54]. IAA was only detected in the plasma of DTHI- and DTI-treated mice (P < 0.005; ~500 to 900 µg) but was absent in mice treated with the blank nanoparticle (Fig. 10A, Supplementary Fig. 6). Importantly, I3A and I3C were only present in DTHI-treated mice (P < 0.005; I3A at nanogram level; I3C at ~20 to 100 µg) (Fig. 10B and C, Supplementary Fig. 6).

Fig. 10.

Detection of reactive IAA metabolites in mice plasma. Plasma concentrations of A. IAA, B. I3A, and C. I3C calculated using Analyst software 1.4.2 from the linear regression equation of the peak area ratio against the concentration of the calibration curve. D. Dose and time dependant effect of I3A on MKN45 cells. E. Dose and time dependant effect of I3C on MKN45 cells. (F). Dose and time dependant effect of I3A on NCI-N87 cells. G. Dose and time dependant effect of I3C on NCI-N87 cells. Data represents the mean±SD (n = 5 or 6), **P < 0.005

Furthermore, we assessed the cytotoxic effects of I3A and I3C on gastric cancer cells. We found that I3A and I3C suppressed the proliferation of GC cells in a dose and time-dependent manner. While IAA alone had no cytotoxic effect (data not shown), the two reactive metabolites had substantial cytotoxic effects at the doses examined (Fig. 10D, E, F and G). Our study provides the first experimental evidence that IAA is converted to cytotoxic metabolites in the presence of HRP and is well absorbed after oral feeding, with plasma concentrations sufficient to elicit biological effects, as demonstrated by the inhibition of MNU-induced gastric tumour growth in mice.

Mass spectrometry analysis of metabolites in mice plasma

To better understand the metabolic mechanisms underlying the protective effect of IAA in the presence of HRP, we then employed a LC-MS/MS-based metabolomic assay and identified metabolites relevant to redox regulations as the top differentially expressed metabolites (alpha-tocopherol [vitamin E] and glutathione [GSH], homocysteine, methyl cysteine, and cysteine sulfinic acid) (Fig. 11 A) [73]. Notably, antioxidants with direct free radical quenching properties (vitamin E and GSH) were upregulated only in the presence of both HRP and IAA, but not in the absence of HRP (Fig. 11B), suggesting a potential advantage of DTHI over DTI in alleviating the carcinogenic effects of MNU, thereby suppressing gastric inflammation and polyp formation (Fig. 9C).

Fig. 11.

Differentially expressed metabolites in plasma between DTI and DTHI treatment as represented with A. volcano plot and B. boxplots. A fold change threshold of 1.5 and a p-value threshold of 0.05 were used to shortlist differentially expressed metabolites. Antioxidants with direct free radical quenching properties are coloured green, and other metabolites relevant to redox regulation are coloured orange. Relative differences in the metabolite levels were analyzed with the Mann-Whitney U-test (n=5)

Discussion

Programmed cell death is a cellular suicide mechanism triggered by the regulated activation of specific cell death pathways and proteins [74, 75]. Necroptosis, a crucial cell death mechanism, is a regulated form of cell death caused by the activation of the necroptotic proteins RIP1, RIP3, and MLKL [29, 30]. While necroptosis has been suggested to increase cancer growth and metastasis, it has also been shown to defend against cancer development independent of apoptosis [76, 78]. Importantly, downregulated expression and functional mutations in the necroptotic pathway regulators have been linked to numerous forms of cancer, including gastric cancer, thus play a key role in escaping cell death [79, 81]. Notably, necroptosis can overcome apoptosis resistance caused by chemotherapeutic drug resistance and elevated levels of anti-apoptotic proteins [82, 83]. It can also induce high adaptive immunity against tumour growth [84, 86]. As a result, the use of therapeutic modulators that can trigger necroptotic cell death has emerged as an effective approach in cancer therapy [87, 88].

We adopted a synergistic approach using BC-based prodrug therapy to specifically sensitize cancer cells to necroptotic cell death. Unlike conventional drug delivery systems, BC-based prodrug therapy enables the localized and controlled generation of therapeutics, minimizing drug delivery and distribution challenges while reducing side effects on healthy tissues. Transition metal catalysts are frequently utilized for bioorthogonal reactions; however, their direct application in the biological environment may be challenging due to their sensitivity to physiological pH fluctuations, resulting in instability and low solubility with limited biocompatibility [89, 90]. Encapsulation technologies which offer exceptional stability, solubility, and efficient cellular uptake, can significantly enhance the in vivo applicability of BC. In this context, we demonstrated the encapsulation of HRP’s coordinated metal reaction centres within the aqueous cavity of exoshells and successfully applied it to the treatment of solid tumours [17]. Prior work from our group established cooperative thermodynamic stability and stoichiometric encapsulation ratios of HRP in consistent with the predicted volume of the exoshells [17, 91]. Furthermore, HRP’s ability to catalyse IAA at low oxygen tensions enables it to target hypoxic tumour areas that are resistant towards necroptosis [54, 80]. DS-tES, a caged nanoparticle engineered to maintain structural integrity under harsh gastrointestinal conditions, was validated by TEM and DLS to remain intact at acidic pH, indicating retention of the encapsulated enzyme (Fig. 1, Supplementary Fig. 1). Although diffusion of HRP through surface pores could be hypothesized, the pore cross-sectional area (~600 Ų) is smaller than the solvent-accessible surface area of native HRP (~820–850 Ų), making diffusion physically impossible [24, 92]. Consistently, SEC analysis of HRP-loaded DS-tES showed enzymatic activity exclusively within cage-containing fractions, with no detectable activity elsewhere (Supplementary Fig. 7), confirming the absence of enzyme leakage. Collectively, these data indicate that neither cage disassembly nor pore-mediated diffusion contributes to enzyme release, supporting the conclusion that DS-tES robustly retains HRP during gastrointestinal transit.

The transit of nanoparticles through the GI tract is often hindered by factors such as harsh pH conditions, proteolytic enzymes, and mucosal and cellular barriers, all of which can compromise the biological activity of the encapsulated therapeutic [93]. We have previously shown that DS-tES effectively stabilized encapsulated Renilla luciferase in the gut for over 24 h, with prolonged residence time likely due to its adhesion to the mucus layer of the GI epithelium [23]. The mucoadhesion may result from non-covalent interactions between the exoshell surface and the mucus layer, or from the formation of disulfide linkages between the exoshells and cysteine-rich subdomains of mucus glycoproteins [94, 95]. Given that surface charge plays a crucial role in mucoadhesion and permeation, we assessed the surface charge of DS-tES under both acidic and basic conditions. Under acidic conditions (pH 4.0), DS-tES exhibited a cationic charge of +14.59 mV, while under basic conditions (pH 7.4), they displayed an anionic charge of −27.52 mV (Fig. 1G). The differential surface charge, which has been widely studied in relation to mucoadhesion and permeation, may influence the permeability of the exoshell [94, 96]. Additionally, DS-tES contain free surface thiols at a concentration of 1.2 micromoles per milligram of protein, which may also contribute to both mucoadhesiveness and cellular uptake via endocytosis, as previously reported [23, 97].

Since DS-tES is a ferritin-based nanoparticle derived from Archaeoglobus fulgidus ferritin, and ferritin nanoparticles are endocytosed through TFRs [98, 99], we investigated the uptake of DS-tES in MKN-45 gastric cancer cells. TFR expression in both NCI-N87 and MKN45 cells was confirmed by western blot analysis (Supplementary Fig. 8) and is consistent with previous reports [100, 101]. To visualize nanoparticle uptake, MKN45 cells were treated with Alexa 488–labeled DS-tES which revealed clear intracellular localization of the nanoparticles (Fig. 8B). Receptor-blocking assays further validated that DS-tES internalization occurs through TFR-mediated pathways (Fig. 8D). Although similar levels of TFR expression were seen on both GC cells, we focused on MKN45 cells and did not assess uptake in NCI-N87 cells. While endocytic uptake of the parent tES nanoparticle has been demonstrated previously [17], we suggest future investigations to determine a full understanding of tissue-level mechanisms governing nanoparticle internalization and barrier traversal.

In this study, we showed that DS-tES encapsulated with HRP, an enzyme centred BC centre, and combined with the prodrug IAA dramatically reduced the tumour volume of gastric cancer through the production of necroptotic modulators. The oxidative decarboxylation mechanism by which HRP catalyses IAA to generate bioactive metabolites, as well as its potential therapeutic application in cancer, has been previously investigated [54, 102]. While our mass-spectrometry analysis quantified I3A and I3C in mouse plasma (Fig. 9B and C), we did not detect other IAA metabolites, such as oxindole-3-carbinol and oxindole-3-acetic acid, which were previously observed in vitro [17]. While both metabolites were found to be cytotoxic to gastric cancer cells (Fig. 9D, E, F and G), only I3C was shown to suppress the proliferation of a variety of gastrointestinal cancers [103, 104]. I3C, a naturally occurring compound derived from IAA metabolism, is known to exert anti-proliferative effects on cancer cells by activating apoptosis related signalling cascades [105, 106]. Beyond their anticancer properties, recent studies in colorectal cancers have shown that I3C mediates cytotoxic and proapoptotic effects via activating the aryl hydrocarbon receptor (AHR) [107]. Likewise, I3A has been identified as an AHR agonist, although its role in causing controlled cell death in cancer cells requires further investigation [108]. Furthermore, both I3A and I3C may enhance gut health by regulating mucosal immunity and intestinal homeostasis via AHR activation [107, 109]. Our differential expression analysis of plasma metabolites revealed post-DTHI elevation in plasma vitamin E and GSH (Fig. 10). Both metabolites are antioxidants with established roles in mitigating inflammation and oxidative stress in damaged tissues [110, 111]. Collectively, these finding highlight the potential role of DTHI-mediated metabolic changes in enhancing anti-cancer activity while mitigating side effects.

Our study is a first of its kind to use a ferritin-based nanoparticle specifically engineered for GI tract stabilization to deliver BC centres via an oral delivery route. Naturally occurring vertebrate and bacterial ferritin’s use as an oral administration agent is severely limited due to an unfavourable gastric environment [112] and subsequent denaturation of the ferritin shell [113]. Further research is also required to thoroughly investigate the safety of inducing necroptosis for anticancer therapy, as induced necroptosis, even when targeted, can cause cell death and inflammation in normal tissues. That said, we have not found any adverse effects to normal tissues following the treatment. We also believe that the technique outlined here, in which necroptotic inducers are generated selectively at target sites, is a promising approach for patients with drug-resistant malignancies [114].

Conclusion

To overcome the challenges associated with delivering bioactive macromolecules to the highly acidic and protease-rich gastric environment, we developed a gastrointestinal tract stabilized iron-mediated catalytic centre encapsulated within the internal cavity of disulfide-linked thermostable exoshells. Our in vivo studies demonstrated that oral administration of these catalytic centres, in conjunction with a prodrug, significantly inhibited gastric tumour growth. This therapeutic effect was mediated through the induction of necroptotic cell death pathways, leading to enhanced antitumor efficacy. This novel strategy offers a promising alternative to conventional systemic infusion therapies, providing a targeted and efficient approach to treating gastrointestinal malignancies.

Author contributions

M.K.S. and G.V.M. contributed equally to this work as co first authors. Conceptualization: M.K.S., G.V.M., S.S., and C.L.D; Methodology: M.K.S., G.V.M., P.Y., H.S.D., L.S.P., L.H.W., L.P.L., and S.S.V.; Data analysis: M.K.S., G.V.M., and C.L.D.; Investigation, M.K.S., G.V.M.,G.S., and C.L.D.; Resources: G.S., and C.L.D.; Writing—original draft preparation: M.K.S., and G.V.M.; writing—review and editing: M.K.S., G.V.M., H.S.D., L.S.P., L.H.W., G.S., and C.L.D.; Supervision: C.L.D.; Project administration: C.L.D.; Funding acquisition: M.K.S., G.V.M., and C.L.D. All authors have read and agreed to the published version of the manuscript.

Funding

This study was supported by fundings from National University Health Science Seed Fund (NUHSRO/2022/063/RO5 + 6/Seed-Mar/08), National Medical Research Council - Clinician Scientist Award - Senior Investigator (MOH-001413-00), and A*Star-MTC-IRG (M23M6c0111).

Data availability

The data that support the findings of this study are available in the manuscript and in the Supplementary Information in the form graphs and figures.

Declarations

Ethics approval and consent to participate

Experiments involving animals were conducted according to the ethical policies and procedures reviewed and approved by the ethics committee of the National University of Singapore, Institutional Animal Care and Use Committee, Singapore (IACUC Protocol No. R22-1119).

Consent for publication

All authors gave their consent for publication.

Competing interests

The authors declare no competing interests.