Synthesis and characterization of Fe1/NC SACsIn our previous study, we proposed a concept of SAzyme and reported an iron-based antioxidative SAzyme35. Since Fe is the active metal center in most natural CATs for catalytic decomposition of H2O2, we aimed to develop Fe-based SACs with similar catalytic functions by mimicking the active site structure of CAT. In addition, Fe is known to exhibit versatile redox chemistry, allowing it to participate in various reduction-oxidation reactions. This redox activity of Fe is crucial for enabling CAT-like activity and other antioxidative activities in scavenging ROS. Fe1/NC SACs were synthesized using an in-situ trapping-pyrolysis method with FePc@MET-6 as precursor at different temperatures, i.e., 800, 900, and 1000 °C, referred to as Fe1/NC-T (T represents the specific pyrolysis temperature) (Fig. 2a). MET-6, composed of Zn clusters and triazolate ligands, was chosen because the volatilization of Zn and decomposition of high-energy triazoles during high-temperature pyrolysis facilitate the generation of gas, forming hierarchical porous carbon network structure that enhances mass transfer and offers accessible active sites for catalysis36. Control samples denoted as NC-T were also synthesized with the same procedure but without FePc. Both MET-6 and FePc@MET-6 show similar octahedral morphology and X-ray diffraction (XRD) peaks (Supplementary Figs. 1 and 2), indicating that FePc does not significantly affect the structure and morphology of MET-6 or the microporous structure (Supplementary Fig. 3 and Supplementary Table 1). However, after high-temperature calcination, all precursors transformed from an octahedral morphology to a hierarchical porous structure. The obtained Fe1/NC SACs show no observable particles in scanning electron microscopy (SEM) and transmission electron microscope (TEM) images (Fig. 2b, c and Supplementary Figs. 4 and 5). Element mapping images demonstrate the uniform distribution of Fe, N, and C in Fe1/NC SACs (Fig. 2d and Supplementary Fig. 6). Aberration-corrected high-angle annular dark-field scanning TEM (AC HAADF-STEM) images show that Fe exists as single atoms for all three Fe1/NC SACs (Fig. 2e–g).Fig. 2: Synthesis and characterizations of Fe1/NC SACs.a Scheme of the synthesis of Fe1/NC SACs. SEM (b) and TEM (c) images of Fe1/NC-900. d HAADF-STEM image and corresponding EDS elemental mapping images of Fe1/NC-900. AC HAADF-STEM images of Fe1/NC-800 (e), Fe1/NC-900 (f) and Fe1/NC-1000 (g). Each experiment was repeated independently three times with similar results. Representative images are shown.In addition, the XRD patterns of Fe1/NC SACs and NC samples show a diffraction peak near 23.3°, corresponding to the (002) crystal plane of graphitic carbon, with no other peaks associated with nanoparticles, further indicating the absence of metal particles in Fe1/NC SACs (Supplementary Fig. 7)37. Raman spectra demonstrate that all the catalysts display two peaks around 1335.7 cm−1 and 1575.4 cm−1, assigned to the D band and G band, respectively (Supplementary Fig. 8)38. Among the NC samples, NC-900 possesses the lowest intensity ratio of ID/IG, which represents the extent of graphitization and is closely related to the calcination temperature, demonstrating the highest degree of graphitization for NC-900 (Supplementary Table 2)39. Fe1/NC SACs show a similar trend with NC catalysts, indicating that the introduction of Fe has little effect on the graphitization of the catalysts. X-ray photoelectron spectroscopy (XPS) and inductively coupled plasma mass spectrometry (ICP-MS) characterizations show similar Fe content in all three Fe1/NC SACs (Supplementary Fig. 9 and Supplementary Tables 3 and 4), while the carbon and nitrogen contents of Fe1/NC SACs depend on the pyrolysis temperature. With increasing the pyrolysis temperature (i.e., 800, 900, and 1000 °C), the carbon content gradually increases (78.18%, 82.07%, and 83.13%, respectively), and the nitrogen content gradually decreases (9.48%, 6.14%, and 5.54%, respectively). In addition, the binding energy of Fe shifts negatively with increasing temperature, indicating a decrease in the valence state of Fe and the attenuation of the interaction between Fe and N (Supplementary Fig. 9c)40. Moreover, the high-resolution N 1s XPS spectra of Fe1/NC can be deconvoluted into three peaks located at around 398.5 eV, 400.0 eV, and 401.1 eV, corresponding to pyridinic nitrogen, pyrrolic nitrogen, and graphitic nitrogen, respectively (Supplementary Fig. 9d)41. Furthermore, the content of pyridinic nitrogen in Fe1/NC SACs gradually decreases with the increase of pyrolysis temperature. These results suggest that the Fe-N coordination environment is highly dependent on the pyrolysis temperature.To further analyze the atomic configuration of the catalysts, we carried out X-ray absorption spectroscopy (XAS) characterization. The Fe K-edge X-ray absorption near-edge structure (XANES) spectra show that the valence states of Fe in Fe1/NC SACs range between 0 (Fe foil) and +3 (Fe2O3) (Fig. 3a), and the energy shifts negetively as the pyrolysis temperature increases, implying a decrease in the oxidation state of Fe, consistent with the XPS results42. Additionally, the extended X-ray absorption fine structure (EXAFS) spectra of Fe1/NC SACs display only one major peak at 1.4 Å in the R space, corresponding to the Fe-N(C) coordination shell, with no Fe-Fe peak similar to that of Fe foil which corresponds to the Fe-Fe coordination shell being observed (Fig. 3b, c), further confirming that Fe in Fe1/NC SACs exists as single atoms. Moreover, the intensity of the Fe-N shell gradually decreases with increasing the pyrolysis temperature, verifying a decrease in the coordination number of N around the Fe center41. To obtain the structural parameters of Fe atoms, we performed EXAFS fitting of Fe1/NC SACs. According to the fitting curves and parameters (Fig. 3d–f and Supplementary Table 5), the coordination numbers of the Fe-N first shell in Fe1/NC-800, Fe1/NC-900, and Fe1/NC-1000 were calculated to be 4.96, 4.09, and 3.26, respectively. In addition to the Fe-N coordination shell, a Fe-C coordination shell with a coordination number of 1.08 was observed in Fe1/NC-1000. Taken together, the atomic configurations of the active centers in Fe1/NC-800, Fe1/NC-900, and Fe1/NC-1000 were depicted as FeN5, FeN4, and FeN3C, respectively, as displayed in the insets of Fig. 3d–f.Fig. 3: XAS spectra of Fe1/NC SACs.Fe K-edge XANES (a) and FT EXAFS (b) spectra of Fe1/NC SACs, Fe foil and Fe2O3. c EXAFS spectra of Fe1/NC SACs, Fe foil and Fe2O3 in k space. Fitting plots of EXAFS spectra for Fe1/NC-800 (d), Fe1/NC-900 (e) and Fe1/NC-1000 (f) in R space. Insets: structural models of Fe1/NC-800, Fe1/NC-900, and Fe1/NC-1000, respectively (N, blue; C, gray; Fe, deep red). Source data are provided with the paper.Antioxidative performance of Fe1/NC SACsHaving demonstrated the coordination environment of Fe active sites can be atomically modulated via high-temperature pyrolysis, we moved forward to investigate the relationship between the CAT-mimicking activity of Fe1/NC SACs and the atomic configuration of their active centers. The CAT-mimicking activity was evaluated by determining O2 generation from the catalytic disproportionation of H2O243,44,45. As shown in Fig. 4a, the as-synthesized Fe1/NC SACs exhibit distinct CAT-like activity, among which, Fe1/NC-900 with the atomic configuration of FeN4 shows the highest activity. In contrast, NC controls without Fe show negligible activity, showing the dominating role of Fe single atoms in mimicking CAT. Moreover, the O2 generation capacities of Fe1/NC SAzymes linearly depend on the concentration of the catalysts (Fig. 4b and Supplementary Fig. 10), implying that the catalytic reaction of the SAzymes follows first order reaction kinetics like natural enzymes46. We next quantitatively determined the specific activities of these three SAzymes, and found that Fe1/NC-900 shows much higher activity (28.9 U·mg−1), nearly three times that of Fe1/NC-800 (10.4 U·mg−1) and Fe1/NC-1000 (11.4 U·mg−1).Fig. 4: CAT-mimicking performance and mechanism investigation of Fe1/NC SACs.a Time-dependent O2 generation in Britton–Robison (BR) buffer containing 5 mM H2O2 and 5 μg·mL−1 catalysts. b Specific activities (SA, U·mg−1) of Fe1/NC SACs in BR buffer (pH 7.0) containing 5 mM H2O2. c Kinetics for CAT-like activity of Fe1/NC SACs (5 μg·mL−1) with different concentrations of H2O2. EPR spectra of 25 mM BMPO in methanol containing 5 mM H2O2 without or with 5 μg·mL−1 Fe1/NC-800 (d), Fe1/NC-900 (e), Fe1/NC-1000 (f) and their controls. Blank: no catalyst. g Schematic structure of Fe1/NC-900 with intermediates during the catalytic H2O2 disproportionation reaction. N, blue; C, gray; O, red; H, light gray; Fe, deep red. h Free energy diagrams of H2O2 disproportionation on Fe1/NC SAzymes. Each experiment was repeated independently three times with similar results (a–f). Representative plots are shown. Source data are provided with the paper.Subsequently, we studied the steady-state kinetics of the three SAzymes in mimicking CAT (Supplementary Fig. 11). As exhibited in Fig. 4c and Supplementary Fig. 12, the decomposition reaction catalyzed by Fe1/NC SAzymes conforms to the typical Michaelis–Menten kinetics. The kinetic parameters of the SAzymes were determined and listed in Supplementary Table 6. Fe1/NC-900 exhibits a maximum reaction rate (Vm) of 0.51 mM/min and a reaction rate constant kcat of 2.05 × 103 min−1, which are much higher than those of Fe1/NC-800 and Fe1/NC-1000 SAzymes, revealing that the highest catalytic activity of Fe1/NC-900. The higher Michaelis constant (Km) value of Fe1/NC-900 indicates a lower binding affinity. In addition, the CAT-like activity of the SAzymes shows an obvious dependence on the pH value of the reaction solution (Supplementary Fig. 13), similar to the natural enzymes47.To evaluate the stability of the Fe1/NC-900 SAzyme, we conducted post-reaction characterizations. As shown in Supplementary Fig. 14a, b, the used Fe1/NC-900 catalyst exhibits a morphology similar to that observed before the catalytic tests. The HAADF-STEM image indicates that no Fe or FexOy nanoparticles were formed in Fe1/NC-900 after catalysis (Supplementary Fig. 14c). The corresponding elemental mapping images confirmed the homogeneous distribution of C, N and Fe elements (Supplementary Fig. 14c). In addition, the XRD spectrum of the used Fe1/NC-900 demonstrates the absence of Fe-related nanoparticles (Supplementary Fig. 15), showing the high stability of the single-atom configuration.In addition to the CAT-like activity, we also explored the SOD-mimicking activity of Fe1/NC-900. SOD is an antioxidative enzyme for catalyzing the disproportionation of O2− into O2 and H2O2. To assess the SOD-like activity of Fe1/NC-900, we employed SOD assay kits with 2-(4-iodophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfophenyl)-2H-tetrazolium (monosodium salt) (WST-1), a selective probe that reacts with O2− to produce a water-soluble formazan dye, to monitor the disproportionation of O2−48,49. The SOD-mimicking ability of Fe1/NC-900 was evaluated and quantified by the inhibition of formazan formation. As shown in Supplementary Fig. 16, the inhibition of formazan formation increases rapidly with increasing the concentration of Fe1/NC-900, indicating that Fe1/NC-900 can mimic SOD and displays a concentration-dependent manner. In comparison, NC-900 without FeN4 active site shows relatively lower SOD-like activity, demonstrating the single Fe atoms in Fe1/NC-900 are responsible for the catalytic disproportionation of O2−.Furthermore, we investigated the activity of Fe1/NC-900 toward eliminating hydroxyl radicals (•OH), another crucial reactive oxygen species. The •OH elimination capacity of Fe1/NC-900 was evaluated and quantified by using terephthalic acid (TA) as a selective fluorescent probe50. TA can be oxidized by •OH to generate 2-hydroxyterephthalic acid (TA-OH), which can emit a fluorescent signal at 435 nm51. As shown in Supplementary Fig. 17, no •OH was detected with TA and H2O2 alone. However, upon the addition of Fe2+, an obvious fluorescence peak representing the TA-OH appears, indicating the generation of •OH through the Fenton reaction. When Fe1/NC-900 was added, the fluorescence intensity of TA-OH significantly decreases, demonstrating that •OH generated from the Fenton reaction can be effectively scavenged by Fe1/NC-900.Given that Fe-based SACs also exhibit peroxidase-like and oxidase-like activities, which may cause cytotoxicity, we systematically evaluated these activities of Fe1/NC-900. We used the oxidation of 3,3′,5,5′-tetramethylbenzidine (TMB) to blue oxidized TMB (ox-TMB) with a characteristic adsorption at ca. 652 nm as a catalytic model reaction. As shown in Supplementary Fig. 18a, Fe1/NC-900 SAC exhibits peroxidase-like activity in BR buffer with a pH value of 4.00. However, under physiological conditions (pH 7.00), Fe1/NC-900 SAC shows negligible peroxidase-like activity, even in the presence of 5 mM H2O2. Similarly, the negligible absorbance of ox-TMB in O2-saturated BR buffer (pH 7.00) indicates that Fe1/NC-900 possesses low oxidase-like activity under a physiological pH (Supplementary Fig. 18b). Therefore, Fe1/NC-900 is unable to cause undesirable oxidative damage due to the low peroxidase-like and oxidase-like activities under the physiological pH.Mechanism study of CAT-mimicking activity of Fe1/NC SACsTo investigate the CAT-mimicking catalytic process of Fe1/NC SAzymes, we conducted electron paramagnetic resonance (EPR) characterization. As shown in Supplementary Fig. 19, the EPR spectra of 5-tert-butoxycarbonyl-5-methyl-1-pyrroline-N-oxide (BMPO) and 2,2,6,6-tetramethyl-4-piperidinone hydrochloride (4-oxo-TEMP·HCl), specific probes for •OH and singlet oxygen (1O2), repectively, show no significant difference with or without Fe1/NC SAzymes, indicating that •OH and 1O2 were not generated during the catalytic process52,53,54. Subsequently, we employed BMPO to capture hydroperoxyl radical (•OOH) in methanol and found that •OOH was generated during the disproportionation reaction of H2O2 catalyzed by the three SAzymes (Fig. 4d–f), which aligns closely with the mechanism of natural enzymes55,56. In addition, the EPR intensity for •OOH correlates well with the CAT-mimicking activity of Fe1/NC SAzymes, indicating that Fe1/NC SAzymes catalyze the decomposition of H2O2 through the generation of the •OOH intermediate.To further elucidate the catalytic mechanism and the difference in CAT-mimicking activity of Fe1/NC SAzymes, we carried out density functional theory (DFT) calculations. Based on the EPR and DFT results, we proposed the reaction pathways for CAT-like activity with Fe1/NC SACs (Fig. 4g), which involve six processes, including a significant •OOH generation step. Whereafter, we calculated the free energies of the reaction steps on the Fe active sites in Fe1/NC to identify the rate-determining step (RDS). As displayed in Fig. 4h and Supplementary Table 7, the adsorption energies of the reactant H2O2 on FeNx (x = 5, 4, and 3) sites in Fe1/NC-800, Fe1/NC-900, and Fe1/NC-1000 are -0.54 eV, 0.11 eV and 0.06 eV, respectively, demonstrating Fe1/NC-800 exhibits the highest binding affinity for H2O2, which is consistent with the result obtained from Km value. Furthermore, by analyzing the Gibbs free energy (ΔG) of the involved steps, we identified the H2O desorption step with a ΔG of 0.41 eV, the H2O2 adsorption step with a ΔG of 0.11 eV, and the •OOH generation step with a ΔG of 0.28 eV as the RDSs for Fe1/NC-800, Fe1/NC-900, and Fe1/NC-1000, respectively. Comparatively, Fe1/NC-900 possesses the lowest ΔG value of RDS, suggesting it has the highest CAT-mimicking activity. These findings demonstrate that the CAT-mimicking activity of SACs can be tuned by modulating the adsorption energy of the reactants and intermediates through atomic engineering of local coordination environment of active sites using high-temperature pyrolysis.Inflammation-free electrochemical sensing with Fe1/NC-900Encouraged by the excellent antioxidative activity of Fe1/NC-900 SAC, we further explored its electrode reactivity toward the oxidation of electroactive neurochemicals. We firstly employed K3Fe(CN)6, a widely used redox probe, to investigate the electrode reactivity of Fe1/NC SACs. Supplementary Fig. 20 shows typical cyclic voltammogram (CV) obtained at Fe1/NC SACs-modified glassy carbon electrodes (GCEs) in artificial cerebrospinal fluid (aCSF) containing 1 mM K3Fe(CN)6. Almost no current ascribed to the redox process of Fe(CN)64-/3- was recorded at Fe1/NC-800-modified GCE, suggesting the poor electrode activity of Fe1/NC-800. In contrast, Fe1/NC-900-modified GCE shows obvious redox peaks with an anodic/cathodic peak-to-peak separation (ΔEp) of about 150 mV, which is smaller than that at Fe1/NC-1000-modified GCE (170 mV), revealing the good electrode activity of Fe1/NC-900.The good electrode activity of Fe1/NC-900 enabled us to develop an in vivo sensing platform for neurochemicals. DA is an important monoamine neurotransmitter that regulates a wide variety of complex neurochemical processes, such as motion, reward, and attention57,58. We chose DA as the target molecule to evaluate the electrochemical sensing ability of Fe1/NC SACs. To do this, we conducted CV measurements of DA at the Fe1/NC SACs-modified GCEs. As shown in Fig. 5a, the electrochemical oxidation of DA at Fe1/NC-900 commences at a potential of ca. −0.05 V (vs. Ag/AgCl) and shows an oxidation peak at ca. + 0.12 V (vs. Ag/AgCl). The onset potential at Fe1/NC-800 was more positive than 0.05 V with a tailed current response (Supplementary Fig. 21a). Fe1/NC-1000 and NC-900-modified GCEs show similar onset potentials compared to the Fe1/NC-900-modified GCE, but with significantly lower oxidation current responses (Supplementary Fig. 21b, c). Moreover, the current responses toward DA oxidation were also recorded with different catalysts under a constant potential (+0.20 V vs. Ag/AgCl). As displayed in Fig. 5b and Supplementary Figs. 22 and 23, the response sensitivity toward DA at Fe1/NC-900 was calculated to be 85 nA/μM, which is much higher than those obtained at Fe1/NC-800 (22 nA/μM), Fe1/NC-1000 (75 nA/μM) and NC-900 (54 nA/μM), suggesting a higher electrode activity of Fe1/NC-900 toward DA oxidation.Fig. 5: In vivo DA sensing with Fe1/NC-900.a CVs obtained at Fe1/NC-900-modified GCE in aCSF in the absence (gray) and presence (red) of 1 mM DA. Scan rate, 50 mV/s. b Plot of current response vs. DA concentration obtained with Fe1/NCs-modified GCEs. Applied potential, +0.20 V vs. Ag/AgCl. c CVs obtained at Fe1/NC-900-modified CFE in aCSF in the absence (gray) and presence (red) of 20 μM DA. Scan rate, 50 mV/s. d Amperometric response recorded at Fe1/NC-900-modified CFE toward successive additions of 5 μM DA in aCSF. Applied potential, +0.20 V vs. Ag/AgCl. Each experiment was repeated independently three times with similar results (a–d). Representative plots are shown. e Amperometric i-t curve recorded with Fe1/NC-900-modified CFE in rat NAc. f Typical amperometric response of Fe1/NC-900-modified CFE in rat NAc upon electrical stimulation of rat VTA (3 s at 60 Hz, ±300 μA) before (gray) and 25 min after (red) the rat were injected i.p. with a DA uptake inhibitor (NOM, 12 mg/kg). Applied potential, 0.20 V vs. Ag/AgCl. The in vivo experiments were repeated with 3 animals. Source data are provided with the paper.Before in vivo analysis, we conducted in vitro experiments on Fe1/NC SACs using SH-SY5Y cell line. The SH-SY5Y cell line was chosen due to its neuronal-like properties, making it a commonly used model system to study various aspects of neurobiology, such as evaluating the effects of neurotoxins and the neuroprotective property of different agents59,60,61. First, we conducted the cytotoxicity experiments of Fe1/NC SACs. As shown in Supplementary Fig. 24, the cell viability assay indicates that the Fe1/NC SACs exhibit minimal cytotoxicity even at a concentration of 50 μg/mL, demonstrating their high biocompatibility. The negligible cytotoxicity of the Fe1/NC SACs suggests their potential application for safe and reliable in vivo neurochemical sensing.We subsequently compared the efficiency of Fe1/NC SACs in protecting cells against oxidative stress using SH-SY5Y cell line. As displayed in Supplementary Fig. 25, H2O2 can induce obvious cytotoxicity on SH-SY5Y cells. Compared with the NC-900 control, the addition of 10 μg mL−1 Fe1/NC SACs to the cell-culture medium effectively attenuates H2O2-mediated oxidative damage and maintains cell viability. Among the SACs tested, Fe1/NC-900 exhibits the highest capability to eliminate H2O2. Furthermore, we used the xanthine oxidation reaction catalyzed by xanthine oxidase (XOD) to produce O2− in situ. After 12-hour exposure to Fe1/NC-900, the damage of SH-SY5Y cells caused by O2− is mostly reduced when compared with the NC-900 control and other Fe1/NC SACs (Supplementary Fig. 26), demonstrating Fe1/NC-900 also possesses the highest SOD-like activity for scavenging O2−. Additionally, we employed the Fenton reaction to produces •OH from the reaction between H2O2 and Fe2+. As shown in Supplementary Fig. 27, Fe1/NC SACs reduce apparent Fenton reagent-induced apoptosis, indicating their specific capability to eliminate •OH. All these results validated that the Fe1/NC-900 exhibits optimal capability to effectively mitigate oxidative stress and protect cells from ROS induced cytotoxicity.Combining the excellent electrode activity with the high antioxidative properties of Fe1/NC-900, we prepared Fe1/NC-900-based carbon fiber electrodes (CFEs) to develop a platform for in vivo DA sensing. As shown in Fig. 5c, d and Supplementary Fig. 28, the Fe1/NC-900-modified CFE exhibits high performance toward DA oxidation with a good linearity between current response and DA concentrations. We implanted the Fe1/NC-900-modified CFE into nucleus accumbens (NAc) to test the stability of the microsensor. As displayed in Fig. 5e, the current response keeps constant over 3600-s measurement, demonstrating the high stability of Fe1/NC-900-based CFE in in vivo sensing. We further used the microsensor to in vivo record DA release in NAc triggered by electrical stimulation of ventral tegmental area (VTA)18. As shown in Fig. 5f, upon electrical stimulation, the current signal ascribed to DA release rises rapidly, reaches its maximal value within few seconds, and quickly decreases to the basal level. To verify the signal comes from DA release, we employed a DA uptake inhibitor, nomifensine (NOM), to block the DA uptake sites, thereby increasing DA overflow62. As illustrated in Fig. 5f, upon intraperitoneal injection of NOM, the current signal significantly increases, showing that the signals detected are indeed attributable to DA release. In addition, we assessed the stability of Fe1/NC-900 under in vivo conditions through TEM characterization. As displayed in Supplementary Fig. 29, the TEM images and corresponding elemental mapping images of Fe1/NC-900 after in vivo experiments show no significant changes, confirming the high stability of the catalyst under physiological environments. These results demonstrate the reliability of Fe1/NC-900 modified CFE for in vivo monitoring of DA dynamics with excellent spatial and temporal resolution.As reported previously, the implantation of CFE into brain tissue inevitably results in a progressive inflammatory tissue response, including the activation of nearby microglia cells and astrocytes, which migrate to the electrode interface. In addition, the increase of hemoglobin can directly upregulate ROS level, further aggravating the inflammatory response15,16,17,18,19. To demonstrate the Fe1/NC-900, with a high antioxidative performance, can potentially endow the microsensor with anti-inflammatory properties, we performed immunohistochemical analysis of brain slices. We used specific markers for activated microglia (ionized calcium-binding adaptor molecule-1, Iba-1) and astrocyte (glial fibrillary acidic protein, GFAP), as well as diamidio-2-phenylindole (DAPI) to stain nuclei63,64.To this end, after the implantation of the microsensor into rat brain for 8 h, brain tissues surrounding Fe1/NC-900-modified and bare (i.e., without surface modification with Fe1/NC-900) CFEs were collected, sliced, and stained for confocal laser scanning microscopy (CLSM) imaging. As displayed in Fig. 6a, Fe1/NC-900 modified CFE elicited a significantly lesser GFAP response toward astrocytes compared to the bare CFE. The markedly reduced intensity and spread of GFAP staining around the Fe1/NC-900 modified CFE indicates a lower level of astrocytic activation, suggesting that the Fe1/NC-900 effectively minimizes the astrocytic inflammatory response (Fig. 6b). Furthermore, more microglia were observed adjacent to the bare CFE compared to the Fe1/NC-900 modified CFE, as indicated by Iba-1 staining (Fig. 6a). The higher density of Iba-1 positive cells around the bare CFE signifies a robust microglial activation and clustering. In contrast, the Fe1/NC-900 modified CFE shows a significantly reduced microglial presence, suggesting that Fe1/NC-900 helps in mitigating microglial activation and the associated inflammatory processes (Fig. 6c). These histological data collectively demonstrate the high efficacy of Fe1/NC-900 in mitigating inflammation, thereby creating an improved microenvironment at neural interfaces adjacent to the CFE.Fig. 6: Histological studies of inflammation in brain tissues.a Histological comparison of brain tissues from untreated rats and rats after 8-hour acute implantation of bare or Fe1/NC-900-modified CFE in NAc (n = 3, for each group). Tissues are labeled for astrocytes (green), microglia (magenta) and nuclei (blue). Scale bar, 50 µm. Fluorescence intensities of astrocytes (b) and microglia (c) in brain tissues from untreated rats (blue line) and rats after 8-h acute implantation of bare CFE (gray line) and Fe1/NC-900-modified CFE (red line) (n = 3, for each group). The distance was measured from the image center for the control group and from the implantation center for the bare and Fe1/NC-900-modified CFE groups. The data were presented as mean ± SEM. Quantitative analysis of ELISA for IL-6 (d), TNF-α (e), and IL-1β (f) in brain tissues from untreated rats (blue column) and rats after 8-hour acute implantation of bare (gray column) and Fe1/NC-900-modified (red column) CFE (n = 3, for each group). The data were presented as mean ± SEM and p values were provided in the figures using one-way analysis of variance (ANOVA). Source data are provided with the paper.In addition, we investigated the expression of typical inflammatory cytokines using enzyme-linked immunosorbent assay (ELISA). Specifically, we measured the levels of interleukin-1 beta (IL-1β), interleukin-6 (IL-6), and tumor necrosis factor-alpha (TNF-α) in the brain tissue surrounding the implant site64,65. The IL-1β level indicates the activation of the inflammatory cascade, while the levels of IL-6 and TNF-α are known to mediate and amplify inflammatory processes. As shown in Fig. 6d–f, the insertion of the bare CFE dramatically upregulated inflammatory cytokines, indicating a pronounced inflammatory response. However, the introduction of Fe1/NC-900 gradually restored the levels of IL-1β, IL-6, and TNF-α to near-normal levels, suggesting an effective suppression of the inflammatory response. This anti-inflammatory effect was attributed to the antioxidative properties of the Fe1/NC-900, which mitigates the oxidative stress and subsequent cytokine production.Overall, Fe1/NC-900 can simultaneously function both as SAzymes for scavenging ROS and as electrode material for DA oxidation, paving an avenue for accurately probing of neurochemical events in vivo with optimized biocompatibility and reduced local acute neuroinflammatory responses.
