Insulin-incubated palladium clusters promote recovery after brain injury

Abstract

Traumatic brain injury (TBI) is a cause of disability and death worldwide, but there are currently no specific treatments for this condition. Release of excess reactive oxygen species (ROS) in the injured brain leads to a series of pathological changes; thus, eliminating ROS could be a potential therapeutic strategy. Herein, we synthesized insulin-incubated ultrasmall palladium (Pd@insulin) clusters via green biomimetic chemistry. The Pd@insulin clusters, which were 3.2 nm in diameter, exhibited marked multiple ROS-scavenging ability testified by the theoretical calculation. Pd@insulin could be rapidly excreted via kidney-urine metabolism and induce negligible adverse effects after a long-time treatment in vivo. In a TBI mouse model, intravenously injected Pd@insulin clusters aggregated in the injured cortex, effectively suppressed excessive ROS production, and significantly rescued motor function, cognition and spatial memory. We found that the positive therapeutic effects of the Pd@insulin clusters were mainly attributed to their ROS-scavenging ability, as they inhibited excessive neuroinflammation, reduced cell apoptosis, and prevented neuronal loss. Therefore, the ability of Pd@insulin clusters to effectively eliminate ROS, as well as their simple structure, easy synthesis, low toxicity, and rapid metabolism may facilitate their clinical translation for TBI treatment.

Keywords: Biomimetic synthesis; Insulin; Palladium cluster; Reactive oxygen species; Traumatic brain injury.

Fig. 1

Scheme illustration. Pd@insulin nanoclusters scavenge excessive multiple ROS, ameliorate a series of negative pathologies, and TBI-induced recovery motor, function, learning, and spatial memory impairment

Fig. 2

Characterizations and multiple ROS scavenging ability of Pd@insulin clusters. A Scheme illustration of Pd@insulin synthesis via the protein biomimetic method. B HRTEM images of Pd@insulin (Scale bar: 10 nm left, 1 nm right). C Superoxide anion E hydroxyl radical, and G H₂O₂ scavenging ability of Pd@insulin (n = 4). Time depended D superoxide anion and F hydroxyl radical scavenging ability of Pd@insulin by EPR spectrum. Comparison of H superoxide anion I hydroxyl radical J H₂O₂ scavenging ability between Pd@insulin cluster, insulin, and Pd²⁺ (n = 5). K Adsorption energies profiles of catalytic process of the SOD and CAT processes for Pd, and L geometry structures of the intermediate states (red and white dots represent oxygen and hydrogen atoms respectively; * represents absorption state). M Intracellular ROS scavenging ability evaluation via N2a, BV2, and A172 cells (Rosup is positive control, green fluorescence represents ROS positive) (Scale bar 200 μm), and N corresponded quantitative results (n = 3). Data are all shown as mean ± SD. Statistical analysis of (C, E, G, H, I, J, and N) was performed by one-way ANOVA with a Tukey post hoc test

Fig. 3

In vitro and in vivo biocompatibility assessment of Pd@insulin clusters. A Blood glucose variation after Pd@insulin/insulin once administration (n = 5, * represents the significant difference between each time point and initial in Pd@insulin group, # represents the significant difference between Pd@insulin and insulin at one time point). B Survival curve of mice after once administrated by Pd@insulin/insulin (n = 5). C Long term blood glucose variation during Pd@insulin daily administration for 6 days (n = 3). D Liver and E renal functions evaluation (n = 3–4). F Representative electrocardiographs of Pd@insulin-treated mice and normal (Ctrl) mice. G Depression and anxiety-like behavior tests of Pd@insulin-treated mice and normal mice (Ctrl) via TST, FST, and SPT (n = 9). Data are all shown as mean ± SD. Statistical analysis of (A) was performed by one-way ANOVA with a Tukey post hoc test, and C–E and G were performed by unpaired Student’s t-test.

Fig. 4

Pd@insulin clusters cross the BBB and aggregate in the injured cortex after TBI. A In vivo blood pharmacokinetic curve. B Excretive Pd²⁺ concentration in mice urine and feces (n = 3). C In vivo fluorescence images of control mouse (left), and mouse treated with Pd@insulin (middle) or insulin (right), respectively. D Biodistribution of Pd@insulin in major tissues of mice at 15 min post treatment. E Fluorescence images of brain tissue post insulin or Pd@insulin administration. F Pd²⁺ concentration in normal mice brain (after perfusion). Immunofluorescent staining of TBI mice for G different brain area including cortex, hippocampus and thalamus. H Immunofluorescent staining of TBI mice cortex (MAP2, Iba1 and GFAP on behalf of neuron, microglia and astrocyte, respectively) (Scale bar 100 μm). Data are all shown as mean ± SD.

Fig. 5

Pd@insulin clusters reverse excessive ROS accumulation, rescue neuronal loss, inhibit microglial activation, and reduce apoptosis after TBI. A Brain ROS level by flow cytometry analyses for Negative, Sham, TBI and TBI mice treated by Pd@insulin mice (10,000 cells per group), and B quantitative analysis (n = 6). C Representative immunofluorescent staining images of NeuN, doublecortin, Sox2, cleaved caspase 3, IBA1, GFAP, TNFα, IL6, BDNF, and TUNEL in injured cortex of TBI mice (Scale bar 100 μm), and D corresponded quantitative analyses (n = 3). Data are all shown as mean ± SD. Statistical analysis of (B, D) was performed by one-way ANOVA with a Tukey post hoc test.

Fig. 6

Pd@insulin clusters ameliorate motor function, cognitive and spatial memory impairment in TBI mice. A Schematic illustration of TBI, treatment and behavior tests. B Balance beam test: step missing ratio on C day 3 D day 7 and E day 14 (n = 10). F Y maze test: G time in novel arm H mean speed and I representative heat map of mice track (n = 10). J Barnes maze test: K latency to target area L percentage of mice that successfully find and enter into target area M representative track (red line) and N entry times in target quadrant O moving distance in target quadrant on test day (n = 10). Data are all shown as mean ± SD. Statistical analysis of (C–E), (G, H), (N, O), and (K) was performed by one-way/two-way ANOVA with a Tukey post hoc test.

Fig. 7

The therapeutic effect of Pd@insulin on TBI occurs via neuroinflammation inhibition. A Venn diagram of the transcriptomic profiles between TBI-Sham and Pd-TBI. B GO enrichment analysis and C KEGG pathway classification of 37 genes, screened out from the overlapped genes between TBI-Sham and TBI + Pd@insulin-TBI. D Protein-to-protein network of 37 genes involved in Pd@insulin treatment. E qRT-PCR analysis of mRNA expression of the genes (n = 3). F Representative immunofluorescent staining images of injured cortex in TBI mice post Cy5-label Pd@insulin injection (Scale bar 100 nm) and G quantitation analysis for the comparison of Pd@insulin intake in different brain cells (n = 3). qRT-PCR analysis of mRNA expression in BV2 treated by Rosup or Pd@insulin + Rosup (n = 3) for H M1/M2 related genes and K the genes from transcriptomics analysis or L related to neuroinflammation. The production of (I TNFα and IL6 in the supernatant of microglia determined by ELISA (n = 3). J Western blot of Rosup/Rosup + Pd@insulin treated microglia. Data are all shown as mean ± SD. Statistical analysis of (E, G, H, I, K, and L) was performed by one-way ANOVA with a Tukey post.