Biological degradation of B/N-doped carbon nano-onions and the autophagy induction in macrophages

Abstract

Boron/nitrogen co-doped carbon nano-onions (BN-CNOs) are an emergent carbon nanomaterial with high structural stability and excellent biocompatibility. They have recently been demonstrated as a viable platform for the targeted delivery of anticancer therapeutics; however, their biodegradation behaviour and impact on immune cells remain unexplored. Here, we investigated the enzymatic and radical-mediated degradation pathways of BN-CNOs, and their surface-oxidised counterpart (oxi-BN-CNOs) using human myeloperoxidase (hMPO), horseradish peroxidase (HRP), and a UV-assisted photo-Fenton (PF) reaction that mimics tumour-associated oxidative environments. The biodegradation progress was tracked using TEM, Raman spectroscopy, XPS, and LC-MS. The results revealed a progressive loss of graphitic order and multilayer disruption, along with the generation of degradation by-products containing oxygen-rich, oxidised aromatic fragments. The degradation efficiency followed the hierarchy: PF > hMPO > HRP; and oxi-BN-CNOs exhibited faster and more extensive degradation than BN-CNOs, consistent with their higher surface oxygenation. Both nanomaterials were non-haemolytic and preserved macrophage metabolic viability, whereas 150 h degradation by-products displayed marked cytotoxicity. In RAW264.7 macrophages, BN-CNOs and oxi-BN-CNOs were efficiently internalised and triggered autophagy, with oxi-BN-CNOs eliciting a quantitatively stronger response. oxi-BN-CNOs markedly increased Beclin-1, ATG5/7/12, and LC3 lipidation while maintaining stable autophagy-gene transcription and minimal inflammatory activation, uncovering a post-transcriptional mode of autophagy regulation. Together, these findings establish oxi-BN-CNOs as degradable, immunologically compatible carbon nano-constructs that selectively enhance autophagy, offering mechanistic insights essential for their continued biomedical translation.

B/N-doped carbon nano-onions could undergo biodegradation in humans and induce autophagy in macrophages.

1. Introduction

Carbon nanomaterials (CNMs) have gained momentum due to their unique properties, such as chemical stability, a high surface area-to-volume ratio, and diverse chemical and electronic properties, which have sparked growing interest across various industrial and biomedical fields.^1,2 They range in size from 1 to 100 nm in at least one principal dimension, and form different structures such as nanowires, films, nanotubes, and nanospheres.³ Carbon nano-onions (CNOs), also known as multi-layered fullerenes, are composed of numerous concentric shells of fullerenes, each <10 nm in size.⁴ Their small size, high surface area, ease of surface modification, low toxicity,⁵ and remarkable in vivo compatibility,^4,6 made CNOs very promising tools for biological applications. The carbon framework of CNOs can accommodate heteroatom incorporation, making them highly suitable for doping.⁷

By co-doping with boron and nitrogen, we have already reported dual-atom (B and N) co-doped CNOs (BN-CNOs), a new member of the CNO family.^7,8 Building on our broader work on CNOs for biomedical applications, we developed BN-CNOs as a complementary nano-onion platform to explore whether heteroatom doping could provide properties advantageous in biological environments. In our previous studies, we demonstrated that BN-CNOs retain the favourable characteristics of CNO-based materials, including good biocompatibility and suitability for further functionalisation. CNOs tend to form aggregates very easily in water, and there have been studies specifically aimed at making them more water-dispersible.^9,10 BN-CNOs exhibit improved aqueous dispersibility comparatively due to new active sites and defects formed on the surface, which is a relevant feature for biological processing and nanomedicine design.^11,12 Furthermore, dopants such as boron and nitrogen may also influence the biocompatibility and biological behaviour of CNOs. Replacing carbon atoms in carbon materials with nitrogen and boron atoms can generate numerous active sites and defects, while also opening the band gap and resembling semiconductor properties.⁸ This modification enhances both adsorption and electrical performance, marking a significant improvement. Recently, we reported the in vitro and in vivo biocompatibility of BN-CNOs and oxidised BN-CNOs (oxi-BN-CNOs), demonstrating the excellent biocompatibility and low cytotoxicity of BN-CNOs.¹² Moreover, the biomedical applications of this new derivative, BN-CNOs, in breast cancer therapy using a pH-dependent drug release mechanism were also recently investigated.¹¹ Although these materials have shown strong potential for biomedical applications, particularly in drug delivery and chemotherapy, understanding their biodegradability under biologically relevant oxidative and immune-mediated conditions is crucial for the clinical translation of BN-CNOs and has not yet been reported.

Although the primary nanoparticles are sub-10 nm in diameter, BN-CNOs and oxi-BN-CNOs are not expected to exist in the studied media as fully individualised particles. Rather, they are present as aggregated colloidal assemblies, typically with sizes above 200 nm in dispersion.¹¹ Accordingly, the biodegradation behaviour discussed here should be interpreted in the context of this realistic colloidal state, in which enzymes and reactive oxygen species interact first with the outer, chemically accessible regions of the aggregates.

Hence, in this work, we study the biodegradation of both BN-CNOs and oxi-BN-CNOs using horseradish peroxidase (HRP), human myeloperoxidase (hMPO), and the UV-assisted photo-Fenton (PF) reaction. The PF treatment specifically mimics a highly oxidative, cancer-relevant environment, where hydroxyl radicals are generated in the presence of Fe²⁺/Fe³⁺ under UV light. This is particularly important, given that BN-CNOs have recently been shown to be a viable nanocarrier for targeted breast cancer therapy.¹¹ Further, we identified the possible degradation by-products of BN-CNOs (from PF reaction) and the in vitro cytotoxicity of by-products in human monocytes (THP1) and mouse macrophage (RAW264.7) cell lines. Finally, we investigated the role of BN-CNOs in inducing autophagy in RAW264.7 macrophage cell lines.

2. Methodology

2.1. Synthesis of pristine BN-CNOs and oxi-BN-CNOs

Boron/nitrogen-doped CNOs were synthesised as previously reported. Briefly, BN-CNOs were obtained through thermal annealing (around 1650 °C) of detonation nano-diamonds (DNDs) in the presence of boric acid under an inert atmosphere.¹¹

2.2. Enzymatic degradation of BN-CNOs and oxi-BN-CNOs by horseradish peroxidase (HRP)

The biodegradation of BN-CNOs was carried out using HRP, similar to our previous work.¹³ Briefly, 84 and 86 µg of BN-CNOs and oxi-BN-CNOs were mixed in PBS and sonicated for 1 min. The dispersed samples were then mixed with 500 µL of 1.1 mg mL⁻¹ HRP, and 2 µL of 50 mM H₂O₂ was added to this complete mixture every 24 h. The reaction was kept for 60 d at RT, and HRP was renewed every 20 d. The control reaction was also performed, where only H₂O₂ was added at 24 h intervals. Furthermore, to analyse the samples, aliquots were collected at different time points (0, 40, and 60 d) and stored at −20 °C.

2.3. Enzymatic degradation of BN-CNOs and oxi-BN-CNOs by human myeloperoxidase (hMPO)

The biodegradation of BN-CNOs and oxi-BN-CNOs was performed using the human myeloperoxidase enzyme (hMPO) derived from human neutrophils (with an activity of 180–220 U mg⁻¹) was commercially bought from Sigma-Aldrich. BN-CNOs and oxi-BN-CNOs (12.5 µg) were mixed with 112.5 µL of 50 mM phosphate buffer (with 140 mM NaCl and 100 µM DTPA). The enzyme hMPO (12.5 µL of 1 µg µL⁻¹) was added to the mixture. Further, 1 µL of 25 mM H₂O₂ was added every hour, to a final concentration of 200 µM H₂O₂. The reaction was performed for 40 h at 37 °C, where hMPO was renewed every 5 h. Control experiments were conducted simultaneously, in which the BN-CNOs were treated with H₂O₂ alone, in the absence of hMPO. To characterise the degradation, aliquots were collected at 0, 25, and 40 h and stored at −20 °C for further analysis.

2.4. Degradation of BN-CNOs and oxi-BN-CNOs by UV-assisted photo-Fenton (PF) reaction

The biodegradation of BN-CNOs and oxi-BN-CNOs was performed via a UV-assisted PF process using a 100 W UV lamp (365 nm, ballasted) as the light source. BN-CNOs and oxi-BN-CNOs (0.31 mL of 0.78 mg mL⁻¹) were dispersed in 2.46 mL of H₂O in the presence of 10 µL of 1 mM FeCl₃·6H₂O in a quartz tube. The pH of the solution was adjusted to 4 using 0.1 M HCl, and 100 µL of 0.05 M H₂O₂ was added. The reaction was performed after the tubes were closed with a septum and placed approximately 10 cm from the UV lamp. The 1 mM FeCl₃·6H₂O was replenished every 35 h, and 100 µL of 0.05 M H₂O₂ was added every 10 h. The total reaction time was 150 h. The control experiments were conducted simultaneously, with only the addition of H₂O₂ in the absence of FeCl₃·6H₂O under UV irradiation for 150 h. For material characterisation, aliquots were collected at 0, 100, and 150 h and stored at −20 °C. The 0 h samples correspond to the untreated starting BN-CNOs and oxi-BN-CNOs and are used throughout the study as the baseline reference for initial material characterisation.

2.5. Raman analyses

The structural changes in BN-CNOs and oxi-BN-CNOs were studied using Raman spectroscopy with a LabRAM HR Raman spectrometer equipped with a 633 nm laser operated at 3% power, and a microscope with a 100× objective. Samples were prepared by drop-casting onto a clean glass slide, followed by drying under IR irradiation. For each sample, ≥5 spectra were recorded at different locations. Further, the average of all plots at different spots was taken, normalised and plotted. The peak deconvolution was performed to better understand the degradation process. First, the normalised plots were baseline corrected. The baseline correction was carried out in Origin 9 software by subtracting the baseline form data, using the baseline mode as ‘user defined.’ Further, the anchor points are defined with 8 points. The baseline correction was carried out the same way for all the plots. Next, the plots were smoothed to 20 points (using adjacent averaging) and were deconvoluted using multiple peak fit with the Lorentz function in Origin 9 software. From the deconvolution results, the full width at half maximum (FWHM), area under the curve for D band (A_D) and area under the curve for G band (A_G) were given, from which A_D/A_G was calculated. I_D/I_G values were measured using baseline-corrected averaged Raman spectra to compare the oxidation of BN-CNOs and oxi-BN-CNOs with the pristine and ox-CNOs for PF-treated samples, which were reported previously.¹³

2.6. TEM analyses

Morphological changes before and after degradation were analysed using TEM. Imaging was performed using an FEI TECHNAI G2 Spirit BioTwin operating at an accelerating voltage of 120 kV, and high-resolution TEM was conducted using an FEI TECHNAI G2 F30 S-Twin operated at 300 kV. Samples were prepared by drop-casting 6 µL of the sample dispersions on a carbon-coated copper grid, followed by drying under IR irradiation for 15 min. In the case where a buffer was used as the medium, prior to the drying process, the grids were washed with Milli-Q water for 30 min to remove the salts.

2.7. Mass spectrometric analyses

To explore possible degradation by-products, mass spectrometric analyses were performed on BN-CNOs and oxi-BN-CNOs samples before and after PF treatment, according to the protocol of Bai, Hao et al.¹⁴ Time-of-flight mass spectrometry (Acquity H) was used for mass spectrometric analysis. Firstly, the BN-CNOs and oxi-BN-CNOs samples were resuspended in a 1 : 1 (v/v) mixture of sample and acetonitrile. This mixture was infused directly into the ESI source at a flow rate of 10 µL min⁻¹. The ESI peak was maintained at 2.57 kV, and the inlet capillary temperature was set at 150 °C. Signals corresponding to decay by-products (present in the 150 h decay spectra, and absent in the 0 h control) were analysed. Data was acquired in positive-ion mode. The signals from the 0 h and the 150 h were compared, and the molecules corresponding to the particular m/z ratio with elements B, N, C, and O were identified in the 150 h samples. Briefly, the mass values which correspond to the molecules were indicated on the plots. The molecules were selected using the software RSC ChemSpider and ChemCalc. The software predicts the molecular structure that could arise from the particular mass value, using elements C, O, N, and B (in this case). Hence, from mass spectrometric analyses, we identified possible degradation products.

2.8. X-ray photoelectron spectroscopy

XPS analyses were performed by drop casting around 6 µL of the 0 and 150 h PF-treated BN-CNOs and oxi-BN-CNOs on a freshly cleaned silicon wafer. Further, it was dried under IR irradiation for 15 min. This process of drops acting and drying was repeated 5 times to obtain a thin layer of samples. Further, analyses were performed using Omicron Nanotechnology XPS. The software CASA-XPS was used to analyse the XPS spectra.

2.9. In vitro cell viability

RAW264.7 mouse macrophage cells (National Centre for Cell Science, NCCS, Pune, India) and THP1 human monocytic cells (National Centre for Cell Science, NCCS, Pune, India) were cultured in Dulbecco's Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin–streptomycin. Cells were maintained at 37 °C in a humidified incubator containing 5% CO₂. After reaching appropriate confluence, the cells were seeded at a density of 1 × 10⁴ cells per well in 96-well plates. BN-CNOs and oxi-BN-CNOs were then introduced to the cells at gradually increasing concentrations (0.005, 0.01, 0.02, and 0.03 mg mL⁻¹). The cells were also exposed to solutions containing BN-CNOs and oxi-BN-CNOs subjected to photo-Fenton degradation for 150 h. Following 24 h of incubation, MTT reagent was added, allowing viable cells to reduce MTT into insoluble formazan crystals. The crystals were dissolved in molecular-grade DMSO, and absorbance was measured at 570 nm using a microplate reader. All experiments were performed in triplicate, and wells containing only cell culture medium served as controls for viability calculations.

2.10. RBC haemolysis assay

Haemolytic activity of BN-CNOs was quantified using a standard red blood cell (RBC) lysis assay. Fresh human whole blood was collected from healthy adult volunteers after obtaining written informed consent under approval from the Institutional Ethics Committee (IEC approval number: IEC-121/2025). All experiments were performed in accordance with the Guidelines of “ICMR National Ethical Guidelines for Biomedical and Health Research Involving Human Participants”, and the study protocol was approved by the Institutional Ethics Committee (IEC) at CSIR-Centre for Cellular and Molecular Biology (CSIR CCMB). A waiver of informed consent was granted by the IEC as the research exclusively utilised leftover/surplus buffy coats, and all samples were irreversibly de-identified and anonymised prior to use. Blood was collected into EDTA-coated tubes and centrifuged at 500 RCF for 5 min to pellet erythrocytes. Plasma and the buffy coat were removed, and erythrocytes were washed three times with phosphate-buffered saline (PBS) before dilution to a final working concentration of 1%. BN-CNO stock solutions were prepared at 20× the desired assay concentration. For each condition, 10 µL of BN-CNO stock was mixed with 190 µL of diluted erythrocytes in U-bottom 96-well plates to obtain final concentrations ranging from 0.78 to 100 µg mL⁻¹. RBCs treated with PBS served as the negative control, while RBCs treated with 1% Triton X-100 served as the positive control for complete haemolysis. Plates were incubated at 37 °C for 1 h, followed by centrifugation at 500 RCF for 5 min. One hundred microlitres of supernatant was transferred to a flat-bottom plate, and absorbance was measured at 540 nm. The percentage of haemolysis was calculated relative to the positive control and expressed as mean ± SD from three independent biological replicates.

2.11. MTT cytotoxicity assay in RAW macrophages

RAW264.7 macrophages (1 × 10⁴ cells per well) were seeded into 96-well plates and allowed to adhere overnight under standard culture conditions (DMEM supplemented with 10% FBS and 1% penicillin–streptomycin). Cells were exposed to increasing concentrations of BN-CNOs or oxi-BN-CNOs (0.78–100 µg mL⁻¹) for 48 h. After treatment, 10 µL of MTT solution (5 mg mL⁻¹) was added, and the mixture was incubated for 4 h at 37 °C to allow mitochondrial reduction of MTT to the insoluble formazan. Crystals were solubilised in 100 µL DMSO, and absorbance was quantified at 570 nm. Viability was normalised to untreated controls and expressed as a percentage.

2.12. Autophagy studies of BN-CNOs and oxi-BN-CNOs

The RAW264.7 macrophage cell line stably expressing GFP-LC3 was generated by Prof. S. Murty Srinivasula's group using the pMYs-IRES-GFP vector, as previously described by Fujita et al., 2011.¹⁵ RAW GFP-LC3 cells (2 × 10⁵) were seeded on coverslips in 24-well plates. Cells were then incubated with BN-CNOs or oxi-BN-CNOs at concentrations of 10 µg mL⁻¹ or 30 µg mL⁻¹ for 12 or 24 h, or maintained as untreated controls for the same time periods. At the indicated time points, coverslips were collected, and cells were fixed with 4% paraformaldehyde. Coverslips were mounted on glass slides using ProLong Gold antifade reagent (P36930, Invitrogen) and allowed to dry before visualisation. Confocal imaging was performed using an Olympus FV3000 laser-scanning microscope with a 60× objective. Z-stack images were acquired with a slice thickness of 0.38 µm. ImageJ software was used for intensity adjustment, and the images were processed to create representative images.

2.13. Autophagy assessment

Autophagy was assessed by monitoring LC3 lipidation in RAW264.7 macrophages. Cells were seeded at a density of 2 × 10⁵ cells per well in 6-well plates and allowed to adhere overnight in complete DMEM containing 10% fetal bovine serum and 1% penicillin–streptomycin. Cells were then treated with oxi-BN-CNOs at concentrations of 50 and 100 µg mL⁻¹ for 48 h. Following treatment, cells were lysed in ice-cold RIPA buffer supplemented with protease inhibitors, and protein concentrations were determined using the bicinchoninic acid (BCA) assay. Equal amounts of protein (25 µg) were resolved by SDS-PAGE and transferred onto PVDF membranes. Primary antibodies against Beclin-1 (D40C5), ATG3 (3415T), ATG5 (D5F5U), ATG7 (D12B11), ATG12 (D88H11), LC3A/B (D3U4C), and β-actin (13E5) were purchased from Cell Signaling Technology (Danvers, MA, USA). Horseradish peroxidase (HRP)-conjugated secondary antibodies were also obtained from Cell Signaling Technology. Immunoreactive bands were detected by enhanced chemiluminescence and quantified using ImageJ. LC3 autophagy activity was determined by calculating the LC3-II/LC3-I ratio. For autophagic flux analysis, 100 nM bafilomycin A1 (BafA1) was added during the final 4 h of treatment, while parallel wells received vehicle controls with or without BafA1. RAW264.7 macrophages (1 × 10⁴ cells per well) were also treated with BN-CNOs or oxi-BN-CNOs at concentrations of 30 and 50 µg mL⁻¹ for 48 h.

2.14. RAW264.7 macrophage treatment and RNA isolation

RAW264.7 macrophages (2 × 10⁵ cells per well) were plated in 6-well plates and incubated overnight. Cells were treated with BN-CNOs or oxi-BN-CNOs at concentrations of 30 and 50 µg mL⁻¹ for 48 h. Following treatment, total RNA was isolated using TRIzol reagent according to the manufacturer's instructions. RNA purity and integrity were confirmed spectrophotometrically and by agarose gel electrophoresis. One microgram of RNA was reverse-transcribed into cDNA using a commercial high-capacity cDNA synthesis kit.

2.15. Quantitative real-time PCR

qPCR was performed using SYBR Green on a real-time PCR platform to quantify transcripts of Beclin-1, ATG5, ATG7, ATG12, and MAP1LC3. GAPDH (Primer sequences, Table S10) served as the endogenous control. Relative gene expression was calculated using the ΔΔCt method and expressed as fold-change relative to untreated macrophages.

2.16. ELISA

To examine cytokine secretion, culture supernatants were collected after 48 h of compound exposure, clarified by centrifugation at 2000 RCF for 5 min, and analysed with a quantitative TNF-α ELISA kit according to the manufacturer's protocols. Absorbance was measured at 450 nm, and cytokine concentrations were interpolated from a standard curve.

2.17. Statistical analysis

Results are presented as mean ± SD of at least three independent biological replicates. Statistical significance was assessed using unpaired Student's t-test (for two-group comparisons) and two-way analysis of variance (ANOVA) followed by Tukey's multiple comparisons test (for multi-group comparisons), relative to the positive control (PC) or cell control as appropriate for each experiment. Significance thresholds were set as follows: *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. All analyses were performed using GraphPad Prism v10.4.1.

3. Results and discussion

BN-CNOs are promising nanomaterials for biomedical applications due to their ease of chemical functionalisation, spherical shape, size and higher biocompatibility compared to CNTs.¹⁶ However, the biodegradability and in vivo fate of BN-CNOs have not yet been reported, which are crucial for clinical translation. Hereby, the biodegradation of BN-CNOs and oxi-BN-CNOs was explored, using human myeloperoxidase (hMPO – secreted by human neutrophils), which generates HOCl in addition to reactive enzyme intermediates in the presence of NaCl and H₂O₂.¹⁷ Moreover, the degradation using the plant enzyme (HRP) and the UV-catalysed photo-Fenton reaction (producing hydroxyl radicals) were also studied, as shown in Fig. 1.

Fig. 1. Schematic representation of different degradation routes of BN-CNOs and oxi-BN-CNOs using the plant enzyme HRP, human enzyme hMPO, and the UV-assisted PF reaction.

3.1. Biodegradation of BN-CNOs and oxi-BN-CNOs using hMPO

First, BN-CNOs were obtained similarly to our previously reported work via thermal annealing of DNDs with boric acid, followed by annealing at 450 °C.^8,11 Further, the oxi-BN-CNOs were produced via chemical oxidation of BN-CNOs by acid treatment to introduce carboxylic functional groups on the surface of the BN-CNOs.⁵ Briefly, the HR-TEM analyses of BN-CNOs and oxi-BN-CNOs (Fig. S1) were performed to confirm multiple concentric circles of CNOs of size in the range of <10 nm and revealed that the doping of B/N or the chemical oxidation didn't cause any morphological changes to their structure. Moreover, the elemental analyses of these materials were performed using XPS (Fig. S2). The BN-CNOs showed 87.2% C 1s, 5.2% O 1s, 4.5% B 1s and 3.1% N 1s. For oxi-BN-CNOs, the O 1s % increased to 9.4%, along with a decrease in B 1s % and N 1s % to 1.6 and 1.8%, respectively, and finally an increased C 1s % of 87.3%. Therefore, these characterisations confirmed the oxidation of BN-CNOs to form oxi-BN-CNOs. The deconvoluted spectra of C 1s, B 1s, N 1s and O 1s of both BN-CNOs and oxi-BN-CNOs were shown in Fig. S3 and S4.

Next, the biodegradability of BN-CNOs and oxi-BN-CNOs was studied by treating hMPO in the presence of H₂O₂ (200 µM) up to 40 h, and hMPO was renewed every 5 h to maintain its activity. The aliquots were collected at different time points to follow the degradation of the CNOs using TEM and Raman spectroscopy. First, HR-TEM analyses were performed on BN-CNOs and oxi-BN-CNOs at 0 and 40 h to examine morphological changes. At 0 h, the concentric graphitic shells of the BN-CNOs (Fig. 2A and B) and oxi-BN-CNOs (Fig. 2E and F) were clearly visible, indicating the multilayer fullerene structure of BN-CNOs.¹³ However, 40 h of treatment with hMPO affected the morphology of both BN-CNOs (Fig. 2C, D, G and H). Notably, the edges of the multilayer CNOs were distinctly visible, which disappeared after 40 h, as observed in Fig. S5 for BN-CNOs. In particular, significant morphological changes were found for both CNOs samples, where the absence of continuous shell-like morphology was observed for both BN-CNOs (Fig. 2D and H). Some additional TEM images were also given in Fig. S6 to completely understand the changes in the morphology of oxi-BN-CNOs after degradation with discontinuous shell like structures. Moreover, the SAED pattern of BN-CNOs at 0 h showed diffraction circles with a very few spots; however, after 40 h, the diffraction signals had weakened drastically. Similar results were also observed for oxi-BN-CNOs, where the diffraction signals were more diffused and poor in the case of 40 h-treated samples compared to the 0 h control. These results indicate that both BN-CNOs and oxi-BN-CNOs had undergone damage via oxidation induced by hMPO treatment for 40 h.

Fig. 2. HR-TEM images of BN-CNOs (A–D) and oxi-BN-CNOs (E–H) after hMPO degradation at 0 and 40 h, respectively, along with the SAED pattern for the respective time point as the inset. The circular black markings showed the point at which SAED measurements were performed. The yellow circular markings correspond to the complete circular multilayers of BN-CNOs and oxi-BN-CNOs at 0 h, which were absent after 40 h.

To corroborate the TEM analyses and understand the chemical and structural changes, Raman spectroscopy was performed before and after degradation. Fig. 3A and B corresponds to the Raman spectra of BN-CNOs and oxi-BN-CNOs, respectively, before and after 40 h of degradation. Before degradation, two sharp peaks were observed corresponding to the characteristic G band (∼1600 cm⁻¹) attributed to the graphitic structure of CNOs with sp² carbon atoms (C C), and the D band (∼1350 cm⁻¹) corresponds to the defect band attributed to the defects or damages to the graphitic structure resulting in sp³ carbon (C–O).¹³ The deconvolution of 0 h and 40 h of BN-CNOs (Fig. S7B and C, respectively) and oxi-BN-CNOs (Fig. S7E and F, respectively) samples were performed. Hence, additionally, a small shoulder peak was observed around 1100 cm⁻¹ for 0 h samples, corresponding to the innermost shell of the carbon nano-onions as per the reported literature.¹⁸ In the case of 40 h degraded BN-CNOs and oxi-BN-CNOs via hMPO, the deconvoluted spectra showed D and G bands, along with other additional defect bands (refer to Table S1, which lists the defect bands and their significance). The band around 1100 cm⁻¹ (corresponding to the innermost shell of CNOs) was more prominent in intensity after 40 h attributing to the fact that more damage must have happened to the surface of the CNOs, thereby exposing the inner shells. Moreover, bands around 1490 cm⁻¹ and 1700 cm⁻¹ were observed, which correspond to the D3 band and C O vibrational modes of carboxylic acids. The FWHM of each deconvoluted peak was given in Table S2. After 40 h of hMPO degradation, the FWHM of the D band have increased in both BN-CNOs and oxi-BN-CNOs, indicating oxidation and destruction to the structure of CNOs. The broadening of the G band was more significant in oxi-BN-CNOs; however, there was no significant change in the FWHM of the G band in BN-CNOs, indicating that more oxidation occurs in oxi-BN-CNOs. The A_D/A_G values for D and G bands have been tabulated (Table S3). The A_D/A_G values for BN-CNOs showed a slight increase from 2.0 ± 0.0 (0 h) to 2.6 ± 0.0 (40 h), and for oxi-BN-CNOs, the increase was negligible from 2.2 ± 0.1 to 2.1 ± 0.1. However, because the bands were significantly broadened after degradation, the comparison of A_D/A_G values might not be appropriate. Therefore, from these results, we could confirm the degradation of BN-CNOs and oxi-BN-CNOs.

Fig. 3. (A) and (B) show Raman analyses of BN-CNOs and oxi-BN-CNOs before and after degradation using hMPO/H₂O₂, respectively. (C) and (D) show the Raman analyses of BN-CNOs and oxi-BN-CNOs before and after degradation using PF degradation. The plot represents the average of 5–10 analyses performed at 5–10 different locations on the sample.

3.2. Biodegradation of BN-CNOs and oxi-BN-CNOs using HRP

Next, the biodegradation of BN-CNOs and oxi-BN-CNOs was studied using the plant peroxidase, HRP, in the presence of H₂O₂ for 60 days. The aliquots were collected at 20-day intervals, and the changes were characterised by TEM and Raman spectroscopy. The morphological changes of BN-CNOs caused by HRP treatment were analysed using TEM. Initially, BN-CNOs at 0 day were found to form large aggregates with distinct edges (Fig. S8A and D); this aggregation was attributed to the CNOs being dispersed in PBS buffer. Further, as degradation progressed, debris became evident by 40 days (Fig. S8B), and the particle edges appeared increasingly diffuse, with no discrete boundaries remaining (Fig. S8E) as observed for the 0 day sample. Further, after 60 days of degradation (Fig. S8C and F), the BN-CNOs exhibited agglomerations, along with disrupted morphology, especially at the edges. Similarly, in the case of oxi-BN-CNOs, TEM analyses showed distinct differences at 0 and 60 days during enzymatic degradation by HRP. Fig. S9A and D showed aggregates of oxi-BN-CNOs at 0 d of HRP degradation. However, after 40 days (Fig. S9B and E), HRP degradation of oxi-BN-CNOs caused sufficient damage to the aggregates with irregular structures formed in between the aggregates and showed more indistinct boundaries, which were more likely to be degraded. Moreover, after 60 days (Fig. S9C and F), the degradation of oxi-BN-CNOs accelerated, resulting in highly disrupted morphology and softened boundaries throughout the aggregates. Additionally, HR-TEM analyses were performed on the 60 day-degraded samples of BN-CNOs and oxi-BN-CNOs. The HR-TEM analyses (Fig. S10A–C) of BN-CNOs and oxi-BN-CNOs (Fig. S10D–F) after 60 days of degradation showed better clarity of the disruption, wherein the discontinuous concentric circles were observed (encircled in Fig. S10). Moreover, the SAED pattern showed no diffraction rings, but rather a diffuse pattern. These results demonstrate that the morphology of BN-CNOs and oxi-BN-CNOs was damaged after 60 days of HRP treatment.

Furthermore, Raman spectroscopy was performed on samples before and after degradation to investigate structural changes in BN-CNOs. Fig. S11A shows the Raman spectroscopic analyses of BN-CNOs at 0, 40 and 60 days. The 0 d spectra of BN-CNOs showed the characteristic D and G bands of CNOs at ∼1350 and ∼1600 cm⁻¹, respectively.¹¹ As degradation progressed, the intensities of the D and G bands decreased, and the bands broadened. After 40 days of treatment, the BN-CNOs exhibited two distinct types of spectra, designated as types 1 and 2. First, the type 1 plot (more similar to 0 h) corresponded to the area with fewer degraded samples. Next, type 2 spectra were assigned to regions where the samples degraded more effectively, as evidenced by decreased intensity and increased broadening of the D and G bands. These two kinds of plots for the same time of HRP treatment could be attributed to the non-uniform degradation of BN-CNOs by HRP. Moreover, at 60 days, there were two kinds of spectra, wherein the type 1 corresponded to the area where there was comparatively less damage compared to type 2 spectra, where less intense D and G bands were found. Similarly, oxi-BN-CNOs (Fig. S11B) exhibited characteristic D and G bands at 0 d. However, after 40 days of HRP treatment, the peaks had significantly broadened, and their intensities had decreased sharply. The deconvolution of 0 day (Fig. S11C and D) and 60 days (Fig. S11E–H) spectra for both BN-CNOs and oxi-BN-CNOs respectively were performed. After degradation, there were many additional defect bands observed both in the case of BN-CNOs and oxi-BN-CNOs, indicating that defects have been introduced significantly on the CNO structure after HRP degradation. The C O band of carboxylic acids were also observed after 60 days of degradation. The FWHM (Table S4) and A_D/A_G (Table S5) values of each deconvoluted peak were tabulated. Therefore, from TEM and Raman analyses, it is evident that oxi-BN-CNOs undergo more degradation compared to the BN-CNOs, which can be attributed to the additional oxygenated groups on oxi-BN-CNOs.

3.3. Degradation of BN-CNOs and oxi-BN-CNOs using the photo-Fenton (PF) reaction

The UV-assisted PF reaction (generating hydroxyl radicals) was conducted for 150 h to elucidate the degradation mechanism and by-products, similar to those observed during the degradation of hexagonal BN sheets.¹⁹ Briefly, both BN-CNOs and oxi-BN-CNOs samples were incubated in the presence of FeCl₃/UV light and H₂O₂ at pH 4 for 150 h, wherein H₂O₂ was replenished every 10 h and FeCl₃ was supplemented every 35 h. The photographs of the vials before and after degradation are shown in Fig. S12. The change in colour of the samples was observed after the PF reaction. Initially, at 0 h, the BN-CNO samples were not dispersible in the aqueous media. However, after degradation via PF reaction for 150 h, the solution turned blackish grey, indicating that oxidation of the material has occurred, making it more dispersible in aqueous media. Similarly, for oxi-BN-CNO, at 0 h, a blackish dispersion was observed due to the greater oxygen functionalities on its surface, making it more aqueously dispersible. However, after 150 h of PF reaction, they turned pale and transparent, indicating material degradation. The aliquots were collected at different time points and were characterised using TEM, Raman spectroscopy, and XPS. The possible by-products were examined using mass spectrometry. First, significant morphological changes were observed for both BN-CNOs and oxi-BN-CNOs under HR-TEM (Fig. S13–S16, respectively). The 0 h TEM analyses clearly showed the concentric spherical morphology of BN-CNOs (Fig. S13B and S14D) and oxi-BN-CNOs (Fig. S15B and S16A) with distinct multilayers. The SAED patterns also showed diffraction rings with faded diffraction spots at 0 h for both BN-CNOs and oxi-BN-CNOs (Fig. S13C and S15C). However, the diffraction pattern disappeared and became diffuse after 150 h of PF reaction, indicating that the multilayer shells were damaged and no longer exhibited polycrystalline behaviour (Fig. S13F and S15F) for BN-CNOs and oxi-BN-CNOs, respectively. This was also supported by the HR-TEM analysis results after 150 h of degradation. The morphology of both BN-CNOs and oxi-BN-CNOs had completely changed, as shown in Fig. S13D, E and S15D, E, respectively. The complete circular morphology of both BN-CNOs (Fig. S14C and F) and oxi-BN-CNOs (Fig. S16C and F) was missing after 150 h of degradation. BN-CNOs exhibited diffused structures with no clear morphology (Fig. S14B and E) and discontinuous concentric circles of oxi-BN-CNOs were observed after degradation, as shown in Fig. S16C and F. Moreover, these HR-TEM images are qualitatively compared in SI to understand the differences in the morphology of BN-CNOs and oxi-BN-CNOs at 0 h and 150 h of degradation.

Further, Raman analyses were performed to understand the structural changes after 150 h of PF reaction on BN-CNOs and oxi-BN-CNOs. The Raman spectroscopic results showed a significant difference between the 0 h and 150 h samples. Fig. 3C and D show the Raman spectra of BN-CNOs and oxi-BN-CNOs, respectively, at different time points (0, 100, and 150 h). At 0 h, the spectra of both BN-CNOs and oxi-BN-CNOs showed distinct D and G bands. After 150 h of PF reaction, BN-CNOs showed a significant decrease in the intensities of both D and G bands (clearly shown in Fig. S17A). Moreover, negligible D and G bands were observed for oxi-BN-CNOs after 150 h of PF treatment (Fig. S17B). Each of the spectra of BN-CNOs and oxi-BN-CNOs for 0 h, 100 h and 150 h was deconvoluted. At 0 h, both BN-CNOs (Fig. S17C) and oxi-BN-CNOs (Fig. S17D) showed highly intense D and G bands. However, after 100 h of degradation, additional bands at ∼1100 cm⁻¹ and D3 bands, corresponding to the inner-shell C band and amorphous C band, respectively, were observed for BN-CNOs (Fig. S17E). In the case of oxi-BN-CNOs (Fig. S17F), there were D4 and D3 bands along with the C O band of carboxylic groups. After 150 h of degradation, both BN-CNOs (Fig. S17G) and oxi-BN-CNOs (Fig. S17H) showed multiple defect bands along with the C O bands. Moreover, the interesting observation was the blue shift of the G band after degradation, which can be attributed to the damage in the CNO graphitic lattice. The FWHM of each deconvoluted peaks were tabulated in Table S6. However, the FWHM of the D and G bands have not been changed significantly in BN-CNOs after degradation. In the case of oxi-BN-CNOs, the FWHM of the D band has increased drastically. However, the same for the G band has decreased after degradation, which could not be significant due to the increased blue shift of the G band after degradation. The A_D/A_G values for both BN-CNOs and oxi-BN-CNOs have been tabulated (Table S7). Moreover, in the case of BN-CNOs, the A_D/A_G ratio initially increased and then decreased from 2.0 ± 0.0 to 2.3 ± 0.1 to 1.7 ± 0.0 (after 0, 100 and 150 h of degradation, respectively). In the case of oxi-BN-CNOs, the A_D/A_G ratio initially increased and then decreased, similar to BN-CNOs, from 1.9 ± 0.0 to 2.2 ± 0.1 (from 0 to 100 h), and at 150 h, the D and G bands were extremely broad, and the A_D/A_G was 1.6 ± 0.1. Since the bands were significantly broadened after degradation, the comparison of A_D/A_G values may not be entirely reliable. Hence, these analyses confirm the degradation of BN-CNOs and oxi-BN-CNOs using the PF reaction.

To support these results, XPS analyses (Fig. S18–S21) were employed to understand the chemical transformation in BN-CNOs and oxi-BN-CNOs after 150 h of degradation via the PF reaction. As the oxidative degradation is initiated at the nanomaterial surface, the XPS-derived surface composition is the most relevant descriptor of the oxidation/degradation of BN-CNOs and oxi-BN-CNOs in this study. The complete comparison of elemental percentages of BN-CNOs and oxi-BN-CNOs before and after degradation was tabulated for better understanding (Table S9). Notably, the % B and % N in PF-treated BN-CNOs reduced from 7.1% B and 6.2% N (0 h, Fig. S18) to trace levels that were not distinguishable from background noise in the survey spectra after degradation; while the % O increased significantly from 5.2% (0 h) to 30.8% (150 h, Fig. S19), indicating very high chemical oxidation of BN-CNOs. Also, from the high-resolution deconvoluted spectra of C 1s, the percentage of carboxyl C has increased from 8.8% to 20.9%, and the C–B peak ∼283 eV has vanished after 150 h. The deconvoluted O 1s spectra also showed an increase in the carboxyl % C–O from 8.2% to 17.8% and the peaks coresponding to B–O or N–O have completely vanished after 150 h. Similarly, for oxi-BN-CNOs, the % B and % N decreased from 2.1% and 1.4% (0 h, Fig. S20), respectively, to levels indistinguishable from background noise after PF treatment; while the % O was increased significantly from 10.7% to 31.8% after 150 h of PF treatment (Fig. S21). From the high resolution deconvoluted spectra of C 1s, the % C–OH, C O and carboxyl (–COOH) was 18.7%, 4.8% and 9.7%, respectively. However, after degradation, the C–O and carboxyl C–O % increased significantly to 40.6% and 19.5%, respectively. Overall, the XPS results clearly demonstrated the drastic oxidation of both BN-CNOs and oxi-BN-CNOs after PF treatment for 150 h, which could be attributed to oxidation in situ-generated hydroxyl radicals with a very high oxidation potential (2.31 V).¹³ These results also supported the TEM and Raman analyses. The generated ·OH radicals (2.31 V) could oxidise the BN-CNOs and oxi-BN-CNOs more prominently than in the case of hMPO, wherein HOCl (1.48 V) was produced along with the enzyme radical intermediates, Compound I (1.16 V) and Compound II (1.34 V).¹⁷ The least ability to degrade BN-CNOs and oxi-BN-CNOs was for HRP, which could be due to its very low redox potential for the enzyme radical intermediates (Compound I and II – low oxidation potential ∼0.9 V), and no reactive oxygen species were produced. These results are consistent with the degradation of other graphene family materials.^1,13

Furthermore, possible degradation byproducts of BN-CNOs were analysed by mass spectrometry. First, the 0 h samples of BN-CNOs and oxi-BN-CNOs (before degradation) showed negligible peaks between m/z 50–300 Da (Fig. S22 and 4, respectively). However, after 150 h of PF treatment, numerous peaks were observed, which could be attributed to possible degradation by-products. The possible by-products were identified by comparing the molecular weight with the combination of the elements present in BN-CNOs and oxi-BN-CNOs, i.e., B, N, C and O, similar to our previous work.^13,20 The possible degradation by-products are mostly oxidised aromatic hydrocarbons, which contain carboxyl or hydroxyl groups, as shown in Fig. 4 (for oxi-BN-CNOs) and Fig. S22 (for BN-CNOs).¹³ In the case of oxi-BN-CNOs (0 h), certain peaks were also observed around 279 Da, which were also present in the 150 h degraded BN-CNO sample. This indicates that the degraded BN-CNOs after 150 h of PF treatment showed a mass spectrum similar to that of oxi-BN-CNOs at 0 h, providing evidence of oxidation of BN-CNOs. However, since the molecules were determined by mass, the isotopes of the proposed molecules are also potential degraded by-product molecules.

Fig. 4. The LC-MS spectra of oxi-BN-CNOs after treating with the PF reaction for 0 (control) and 150 h, and the possible degradation by-products after 150 h.

3.4. Comparison of the degradation of BN-CNOs with other carbon materials

CNOs exhibited notably higher degradation resistance compared to other graphitic materials, such as graphene oxide (GO)²⁰ and single-walled carbon nanotubes (SWCNTs),²¹ under oxidative conditions, using oxidative enzymes such as hMPO secreted by neutrophils. This was evidenced by slower structural damage and less incorporation of oxygen groups post-degradation. GO took around 24 h to show nearly complete degradation using hMPO; SWCNTs were also reported to degrade by 24 hours of hMPO treatment.²¹ However, the CNOs took 40 h for partial degradation, which could be attributed to their multiple layered structure.¹³ Moreover, when degradation via PF reaction was compared, GO required around 72 h for degradation, whereas 150 h were required to partially degrade CNOs¹⁴ similar to multiwall CNTs (MWCNTs), which took 168 h to degrade 95% of the structure.²² This can be correlated to the similar structure of MWCNTs and CNOs in terms of multiple layers, wherein the former is tube-like and the latter is spherical.

Further, it is relevant to compare hexagonal boron nitride (hBN), which is an analogue of graphene known as white graphene with graphitic materials such as GO and CNTs, in order to better understand the effect of B and N doping. hBN was reported to exhibit greater oxidation resistance than GO and CNTs.¹⁹ hBN took around 100 h of PF reaction and 35 h of hMPO treatment to show significant degradation and highly porous structures, unlike GO and SWCNTs. Additionally, Kostoglou et al. reported that graphene underwent thermal oxidation at 450 °C, while the BN monolayer was oxidised only at 850 °C, making them thermally more resistant to oxidation.²³ Hence, from these findings, it is evident that the chemical structure of hBN, with strong B–N bonds, and the multilayered structure of CNOs and MWCNTs are the reasons for the clear oxidation resistance compared to other graphitic materials like GO, SWCNTs, etc. Therefore, these studies support that BN-CNOs are proposed to have higher stability towards degradation.

Next, when pristine CNOs (p-CNOs) and BN-CNOs were compared, the degradation of p-CNOs¹³ was quite similar to BN-CNOs after 150 h of UV-catalysed PF reaction, with slightly higher degradation resistance observed in BN-CNOs compared to p-CNOs. This was indicated by the XPS analyses data (Table 1), which revealed that the O 1s content increased to 37.6% in p-CNOs but only 30.8% in BN-CNOs post-degradation. Also from the Raman analyses (Table 2), the I_D/I_G increased from 1.55 (0 h) to 1.99 (100 h) and the D and G bands were negligible after 150 h for p-CNOs after PF reaction, whereas in BN-CNOs, the I_D/I_G values increased from 1.64 (0 h) to 1.75 (100 h) and decreased to 1.41 after 150 h PF treatment. These results indicated that B/N doping might have introduced shields to oxidise the BN-CNO surface, even though defects were present on it due to doping. Therefore, from the literature support and the experimental results, a slightly higher oxidation/degradation was observed for p-CNOs compared to BN-CNOs, although these differences were not highly significant.

Table 1. Shows the atomic percentages of p-CNOs, oxi-CNOs, BN-CNOs and oxi-BN-CNOs at different time points of PF reaction.

Atomic %	Time	C 1s %	O 1s %	B 1s %	N 1s %
p-CNOs13	0 h	99	1	—	—
	150 h	62.44	37.56	—	—
oxi-CNOs13	0 h	90.7	9.3	—	—
	150 h	60.72	39.28	—	—
BN-CNOs	0 h	81.43	5.24	7.09	6.24
	150 h	69.22	30.78	—	—
oxi-BN-CNOs	0 h	85.76	10.74	2.14	1.36
	150 h	68.19	31.81	—	—

Table 2. Shows the I_D/I_G values of p-CNOs, oxi-CNOs, BN-CNOs and oxi-BN-CNOs at different time points of PF reaction.

I D/IG	0 h	100 h	150 h
p-CNOs13	1.55	1.99	—
oxi-CNOs13	1.34	2.39	—
BN-CNOs	1.64	1.75	1.41
oxi-BN-CNOs	1.45	1.48	1.29

3.5. BN-CNO formulations display excellent viability along with haemocompatibility

THP1 monocytes retained high metabolic viability following exposure to increasing doses of BN-CNOs and oxi-BN-CNOs, indicating that both formulations are well tolerated across the tested concentration range (Fig. 5A). Even at the highest doses, viability remained comparable to untreated controls, confirming that the compounds do not elicit cytotoxic stress capable of confounding downstream signalling or autophagy measurements. Notably, oxi-BN-CNOs preserved macrophage viability with minimal fluctuations across doses, supporting the interpretation that subsequent changes in autophagy markers arise from genuine biological modulation rather than loss of cell fitness. Even at the highest concentration tested (100 µg mL⁻¹), viability remained consistently high (>85–90%), with absorbance values deviating by <10–15% from those of untreated controls. The oxi-BN-CNOs showed a slightly more favourable viability profile, with viability remaining nearly superimposable to that of controls across all doses, indicating that the compound does not exert overt cytotoxicity and permitting a clean biochemical interpretation of its downstream signalling effects. Surprisingly, BN-CNOs and oxi-BN-CNOs exposed to PF degradation showed a pronounced reduction in metabolic viability at 150 h, indicating that the long-term degradation by-products exert measurable cytotoxic effects on macrophages (Fig. 5B). BN-CNOs and oxi-BN-CNOs exhibited minimal hemolytic activity across all tested concentrations, 100–0.78 µg mL⁻¹ (Fig. 5C). Having established that BN-CNOs and oxi-BN-CNOs do not compromise macrophage viability, we next evaluated their compatibility with erythrocytes to ensure that the materials do not disrupt membrane integrity under physiological conditions. Haemolysis levels remained comparable to the PBS control and far below the Triton X-100 PC, indicating that both BN-CNOs and oxi-BN-CNOs do not disrupt erythrocyte membrane integrity under physiologically relevant exposure conditions.

Fig. 5. (A) Cell viability analysis by MTT. THP1 monocytes were treated with serial dilutions of BN-CNOs or oxi-BN-CNOs for 48 h. Metabolic activity was assessed by MTT reduction, and values were normalised to untreated controls. (B) Cell viability analysis by MTT in RAW264.7 macrophages at 5, 10, 20, 30 µg mL⁻¹. Statistical significance was compared relative to PC. (C) Haemolytic activity of BN-CNOs and oxi-BN-CNOs. Human erythrocytes were exposed to increasing concentrations of each compound (0.78–100 µg mL⁻¹) for 1 h at 37 °C, and haemoglobin release was quantified as a measure of membrane disruption. The percentage of haemolysis was calculated relative to 1% Triton X-100 (PC). Data represent mean ± SD (n = 3). Statistical significance was determined by two-way ANOVA with Tukey's multiple comparison test (*p < 0.05; **p < 0.01; ***p < 0.001; and ****p < 0.0001).

3.6. In vitro cytotoxicity and autophagy studies of BN-CNOs and oxi-BN-CNOs

Many carbon nanomaterials are known to interact with and be internalised by macrophages. Several studies have also shown that nanomaterials can activate immune cells and induce autophagy.²⁴ Microtubule-associated protein 1A/1B-light chain 3 (LC3) is a well-established marker of autophagy.²⁵ Upon induction of autophagy, cells initiate the formation of autophagosomes. During autophagosome maturation, LC3 undergoes lipidation with phosphatidylethanolamine, converting the cytosolic LC3-I form into LC3-II, which incorporates into the autophagosomal membrane. This is an essential step in autophagosome biogenesis.²⁶ Because BN-CNOs undergo enzymatic degradation and are efficiently internalised by macrophages, we next examined whether their uptake modulates cellular homeostasis through autophagy, a well-recognised pathway engaged during nanoparticle processing (Fig. S23 and S24). This analysis also provides a mechanistic link between biocompatibility, intracellular trafficking, and the biochemical effects observed in later autophagy marker assays.

To examine whether CNOs can interact with immune cells and induce autophagy, RAW264.7 macrophages stably expressing GFP-LC3 were incubated with BN-CNOs or oxi-BN-CNOs. After 12 h, both nanomaterials were visibly internalised by GFP-LC3 RAW264.7 cells, and their uptake appeared dose-dependent. Concurrently, cells internalising BN-CNOs or oxi-BN-CNOs displayed prominent punctate GFP-LC3 positive structures compared to untreated controls, indicating autophagosome formation and the initiation of autophagy at 24 h. These observations suggest that both BN-CNOs and oxi-BN-CNOs trigger autophagy in RAW264.7 macrophages. Interestingly, oxi-BN-CNOs appeared to be taken up more efficiently than BN-CNOs, suggesting enhanced biological interaction of the oxidised form. However, this observation was further validated.

3.7. oxi-BN-CNOs robustly increase autophagy marker abundance and drive efficient LC3 lipidation despite stable autophagy gene expression

Western blot analysis reinforced quantitatively superior protein-level response to oxi-BN-CNOs. Across all autophagy markers examined, oxi-BN-CNOs produced stronger induction than BN-CNOs (Fig. 6A). Beclin-1 levels increased by roughly 1.6–2.0-fold with oxi-BN-CNOs (50 µg mL⁻¹), compared with a modest 1.1–1.3-fold shift with BN-CNOs. ATG5, ATG7, and ATG12 all showed similar magnitudes of enhancement, with oxi-BN-CNOs yielding mean increases in the 1.5–2.2-fold range depending on the marker and dose, whereas BN-CNOs repeatedly trailed with <1.4-fold induction. The most functionally consequential effect was observed in LC3 processing. oxi-BN-CNOs significantly increased LC3-II levels while decreasing or maintaining LC3-I, resulting in a pronounced elevation in the LC3-II/LC3-I ratio. At 50 µg mL⁻¹, oxi-BN-CNOs produced an increase in LC3 lipidation approaching 2.5–3-fold over control, substantially greater than the 1.2–1.5-fold shift observed with BN-CNOs (Fig. 6B). Even at 30 µg mL⁻¹, oxi-BN-CNOs sustained LC3-II/LC3-I ratios well above baseline, indicating that autophagosome maturation is strongly engaged even at submaximal dosing.

Fig. 6. Modulation of autophagy signalling and inflammatory responses in RAW264.7 macrophages following BN-CNO and oxi-BN-CNO treatment. (A) Immunoblot analysis of autophagy markers. RAW264.7 cells were treated with BN-CNOs or oxi-BN-CNOs (30 and 50 µg mL⁻¹, 48 h), and whole-cell lysates were examined for Beclin-1, ATG3, ATG5, ATG7, ATG12, and LC3A/B. β-Actin served as the loading control. Representative blots from three biological replicates are shown. (B) Densitometric quantification of autophagy proteins. Protein band intensities from panel (A) were normalised to β-actin and expressed relative to untreated control cells. The LC3-II/LC3-I ratio was used as a surrogate for autophagosome maturation. Data are presented as mean ± SD (n = 2). (C) Transcriptional profiling of autophagy-related genes. Relative mRNA expression of Beclin-1, ATG5, ATG12, ATG7, and MAP1LC3 in RAW264.7 cells treated with BN-CNOs or oxi-BN-CNOs (30 or 50 µg mL⁻¹, 48 h). Transcript abundance was normalised to GAPDH and expressed as a fold change over control. Data represent mean ± SD (n = 3). (D) TNF-α secretion following compound treatment. Supernatants collected from RAW264.7 macrophages after 48 h of exposure to BN-CNOs or oxi-BN-CNOs were analysed by ELISA. (E) LC3 and β-actin immunoblot analysis illustrating autophagic flux across oxi-BN-CNOs and BafA1 treatment conditions, with densitometric bar graph. Cytokine concentrations were normalised to untreated controls and are shown as mean ± SD (n = 3). Statistical significance was determined by one-way ANOVA with Tukey's multiple comparison test (*p < 0.05; **p < 0.01; ***p < 0.001).

oxi-BN-CNOs maintained autophagy-related (ATG) gene expression within a narrow homeostatic range while simultaneously driving strong autophagic protein accumulation (Fig. 6C). Beclin-1 mRNA remained close to baseline under oxi-BN-CNOs treatment, measuring 1.0-fold at 50 µg mL⁻¹ and 0.85-fold at 30 µg mL⁻¹, in contrast to LPS, which reduced Beclin-1 to 0.70-fold. ATG5 expression was unchanged across all groups (0.9–1.2-fold), while ATG12 remained near basal values following oxi-BN-CNOs (0.90-fold at 50 µg mL⁻¹; 0.75-fold at 30 µg mL⁻¹), sharply contrasting the LPS-driven surge to 2.1-fold. ATG7 expression, which was markedly suppressed by LPS (0.55-fold), was fully restored by oxi-BN-CNOs to approximately 1.10-fold, matching the levels of untreated controls. MAP1LC3 likewise showed no induction under oxi-BN-CNOs, in contrast to the pronounced LPS peak. Despite this transcriptional stability, oxi-BN-CNOs induced a robust elevation of autophagy proteins, including Beclin-1, ATG5, ATG7, and ATG12, and produced the strongest LC3 lipidation response, reflected by a markedly increased LC3-II band and a significant rise in the LC3-II/LC3-I ratio. In contrast to BN-CNOs, which elicited substantial TNF-α secretion, oxi-BN-CNOs maintained a low-inflammatory signature while strongly promoting autophagy (Fig. 6D). Together, these findings reveal that oxi-BN-CNOs act primarily at the protein level, stabilising the autophagy machinery and enhancing autophagosome maturation without requiring transcriptional induction.

3.8. oxi-BN-CNOs trigger a comprehensive autophagic response

Taken together, the dataset reveals that oxi-BN-CNOs exert a markedly stronger autophagy-inducing phenotype than BN-CNOs across transcriptional, translational, and functional biochemical metrics. The compound consistently drives 1.5–3-fold elevations in key autophagy determinants, substantially enhances LC3 lipidation, and does so while maintaining high cellular viability and minimal inflammatory skewing. The convergence of these quantitative signatures establishes oxi-BN-CNOs as a chemically distinct macrophage modulator that preferentially amplifies autophagy, positioning it as a superior candidate for applications requiring selective activation of autophagy without concomitant inflammatory activation. To determine whether the observed increases in LC3 puncta and LC3-II abundance reflected true autophagic flux rather than impaired degradation, we next performed a lysosomal inhibition assay using bafilomycin A1 to quantify LC3-II accumulation in the presence and absence of oxi-BN-CNOs (Fig. 6E). Densitometric assessment of LC3 immunoblots showed that autophagic flux was markedly enhanced under combined BafA1 + oxi-BN-CNO treatment. While oxi-BN-CNOs alone induced a moderate increase in LC3-II relative to untreated control conditions, the addition of BafA1 resulted in a pronounced accumulation of LC3-II, reflecting inhibited lysosomal degradation and revealing higher autophagosome turnover. The stronger LC3-II band intensity in the BafA1 + oxi-BN-CNO lanes, compared with oxi-BN-CNOs alone, indicates that oxi-BN-CNOs elevate autophagic activity and that this flux becomes unmasked upon lysosomal blockade. β-Actin levels remained stable across all samples. Uncropped immunoblots are available in the SI (Fig. S25).

Unlike many canonical autophagy activators that rely on transcriptional reprogramming or metabolic stress, oxi-BN-CNOs enhance autophagic activity while maintaining stable expression of core autophagy genes. Across all targets assessed, including Beclin-1, ATG5, ATG12, ATG7, and MAP1LC3, oxi-BN-CNOs preserved mRNA levels near baseline, even under conditions where LPS profoundly disrupted transcriptional patterns. The maintenance of transcriptional homeostasis suggests that oxi-BN-CNOs do not initiate autophagy through classical gene-induction pathways, but rather act through post-transcriptional or post-translational mechanisms. Despite minimal changes at the mRNA level, oxi-BN-CNOs elicited a robust augmentation of autophagy at the protein level. The compound significantly increased the abundance of Beclin-1, ATG5, ATG7, and ATG12, producing the strongest LC3 lipidation response among all tested treatments. The substantial elevation in LC3-II and the pronounced increase in the LC3-II/LC3-I ratio indicate efficient autophagosome maturation, suggesting that oxi-BN-CNOs enhance autophagy through modulation of the conjugation machinery, vesicular dynamics, or upstream kinases that control LC3 processing. The divergence between stable transcription and elevated protein levels underscores a post-transcriptional mode of regulation, potentially involving enhanced protein stability or altered autophagosomal turnover.

Importantly, although BN-CNOs and oxi-BN-CNOs induced cytokine production compared with the untreated control, the response remained indicative of a controlled immunomodulatory profile rather than a broad pro-inflammatory response. This inflammation-sparing profile is particularly advantageous, as autophagy itself limits inflammatory stress by reducing mitochondrial ROS and modulating inflammasome activity, suggesting autophagy activation without inducing excessive inflammatory signalling.

The increase in LC3-II upon lysosomal inhibition confirms the presence of ongoing autophagic flux rather than a static buildup of autophagosomes. In other words, the system actively forms and degrades autophagosomes, and the lysosomal block unmasks this dynamic process. Such LC3-II accumulation is a classic hallmark of functional autophagy, suggesting that the tested condition modulates autophagic throughput rather than simply altering LC3 expression. This pattern aligns with mechanistic models where stress, metabolic rewiring, or targeted compounds enhance autophagosome biogenesis, and flux readouts distinguish true induction from impaired degradation.

To further contextualise the macrophage response, we compared the behaviour of BN-CNOs and oxi-BN-CNOs with that of previously reported carbon nanomaterials. Graphene oxide (GO), for instance, is well known to undergo rapid enzymatic degradation (within ∼24 h under hMPO conditions) and to induce pronounced oxidative stress and inflammatory activation in macrophages.^27,28 In contrast, BN-CNOs exhibit comparatively slower degradation kinetics and, importantly, do not elicit strong inflammatory cytokine responses under the conditions tested. This divergence is likely attributable to the multiple concentric layers of CNOs, which make them more difficult to degrade than single-layered GO sheets. Moreover, the structural stability conferred by B/N co-doping and the distinct surface chemistry of nano-onion architectures also contribute. Furthermore, boron nitride-based nanomaterials are generally regarded as more chemically inert and biocompatible than graphene derivatives, supporting the minimal TNF-α secretion observed in our system. Within this framework, oxi-BN-CNOs display a unique immunobiological profile characterised by robust autophagy induction coupled with low inflammatory activation, distinguishing them from conventional carbon nanomaterials, which often trigger stress-associated or pro-inflammatory pathways. While a direct side-by-side experimental comparison with GO or hBN was beyond the scope of this study, these observations position oxi-BN-CNOs as a distinct class of immunomodulatory nanomaterials with favourable macrophage compatibility.

Collectively, these findings position oxi-BN-CNOs as a non-cytotoxic, low-inflammatory, and post-transcriptionally acting autophagy enhancer with potential applications in host-directed antimicrobial strategies, macrophage immunomodulation, and vaccine adjuvant design. Further investigation into its upstream molecular targets and in vivo activity will clarify its translational value.

4. Conclusion

The biodegradation of BN-CNOs and oxi-BN-CNOs by the plant enzyme HRP, the human peroxidase enzyme hMPO, and the PF reaction was studied. In particular, biodegradation was likely initiated at the outer, chemically accessible regions of aggregated CNO assemblies and proceeded nonuniformly, with inner graphitic shells affected only progressively over time. This is why gradual changes were observed, including limited spectral differences at intermediate time points and pronounced degradation at later stages. The oxi-BN-CNOs showed greater degradability than BN-CNOs, which could be attributed to the presence of a large number of oxygenated groups that could serve as initial sites for oxygenation or degradation. Moreover, we observed that degradation was maximum in the PF reaction, followed by hMPO and, finally, by HRP. This trend is also evident in other graphitic materials.^13,24 This might be due to the high redox potential of the reactive oxygen species produced in each type of degradation. Moreover, in biological studies, the oxi-BN-CNOs were found to be non-cytotoxic, to elicit a low inflammatory response, and to act as an autophagy enhancer. Overall, these findings provide important insight into the biodegradation behaviour and macrophage response of BN-CNOs and oxi-BN-CNOs, thereby helping to define their future design and translation for biomedical applications.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting this article have been included in the main text and supplementary information (SI). Supplementary information: experimental procedure and additional figures, cited additional references. See DOI: https://doi.org/10.1039/d6ra04104c.