Core-Cross-Linked Nanoparticles Reduce Neuroinflammation and Improve Outcome in a Mouse Model of Traumatic Brain Injury

Abstract

Traumatic brain injury (TBI) is the leading cause of death and disability in children and young adults, yet there are currently no treatments available that prevent the secondary spread of damage beyond the initial insult. The chronic progression of this secondary injury is in part caused by the release of reactive oxygen species (ROS) into surrounding normal brain. Thus, treatments that can enter the brain and reduce the spread of ROS should improve outcome from TBI. Here a highly versatile, reproducible, and scalable method to synthesize core-cross-linked nanoparticles (NPs) from polysorbate 80 (PS80) using a combination of thiol-ene and thiol-Michael chemistry is described. The resultant NPs consist of a ROS-reactive thioether cross-linked core stabilized in aqueous solution by hydroxy-functional oligoethylene oxide segments. These NPs show narrow molecular weight distributions and have a high proportion of thioether units that reduce local levels of ROS. In a controlled cortical impact mouse model of TBI, the NPs are able to rapidly accumulate and be retained in damaged brain as visualized through fluorescence imaging, reduce neuroinflammation and the secondary spread of injury as determined through magnetic resonance imaging and histopathology, and improve functional outcome as determined through behavioral analyses. Our findings provide strong evidence that these NPs may, upon further development and testing, provide a useful strategy to help improve the outcome of patients following a TBI.

Keywords: antioxidant; controlled cortical impact; gliosis; hippocampus; polysorbate 80; startle habituation.

Figure 1.

(A) ¹H NMR of PS803SH in CDCl₃ following thiol–ene conjugation of tetra-SH to PS80. (B) Overlay of the olefin resonance of PS80 before (red trace) and after (black trace) thiol conjugation confirming quantitative reaction of the cis double bond. (C) ¹H NMR spectrum in CDCl₃ for core-cross-linked NP1 prepared with the addition of pentaerythritol tetraacrylate (TAc) to an aqueous solution of PS803SH and at an overall olefin to thiol ratio of 1. (D) GPC chromatograms for PS80 and core-cross-linked NPs prepared from the reaction of PS803SH (with (F, G) or without (E) addition of free thiol) and TAc at an overall thiol to olefin ratio of 1. Absolute molecular weights and Ð values were determined via GPC in dimethylformamide eluent at a flow rate of 1 mL/min.

Figure 2.

(A) Schematic representation of therapeutic NPs scavenging reactive oxygen species release following traumatic brain injury. (B) DCFH fluorescence as a function of H₂O₂ concentration with and without the addition of NP1. In the absence of ROS-scavenging NPs a significant increase in fluorescence is observed above an H₂O₂ concentration of 1 mM. In contrast, experiments conducted in the presence of NP1 show significant reduction in fluorescence even at low concentration.

Figure 3.

NP1 reduced ROS-mediated activation of human astrocytes as determined by GFAP immunostaining. (A) Media (10×), (B) 1 mg/mL NP (10×), (C) 0.1 mg/mL NP + 100 μM H₂O₂ (10×), (D) 100 μM H₂O₂ (10×), (E) media (60×), (F) 1 mg/mL NP (60×), (G) 0.1 mg/mL NP + 100 μM H₂O₂ (60×), (H) 100 μM H₂O₂ (40×). (I) Quantification of numbers of GFAP+ cells in low-powered field images as compared to total cell number determined by DAPI nucleus staining. There was a significant difference in the percentage of GFAP+ cells in vitro between treatments (P = 0.0002), with NP treatments significantly reducing the percentage of GFAP+ cells back to baseline as compared to H₂O₂ treatment (adjusted P values are media-only vs NP-only adjusted, 0.8797; media-only vs H₂O₂, 0.0258; media-only vs 1 mg/mL NP + H₂O₂, 0.6865; media-only vs 0.01 mg/mL NP + H₂O₂, 0.2424; NP-only vs H₂O₂, 0.0027; NP-only vs 1 mg/mL NP + H₂O₂, 0.9955; NP-only vs 0.01 mg/mL NP + H₂O₂, 0.7597; H₂O₂ vs 1 mg/mL NP + H₂O₂, 0.0011; H₂O₂ vs 0.01 mg/mL NP + H₂O₂, 0.0001; 1 mg/mL NP + H₂O₂ vs 0.01 mg/mL NP + H₂O₂, 0.9251).

Figure 4.

(A) NP1 accumulates in damaged brain after intravenous injection. Fluorescence images show high NP1 accumulation quickly after injection, which was retained in damaged brain for over 2 h postinjection. (B) Biodistribution and clearance of NP1. Fluorescence images and quantification show NP1 mainly in the liver and kidneys. Organ layout: 1. liver, 2. kidneys, 3. spleen, 4. heart, 5. lungs. (C) Time-dependent biodistribution of NP1 showing clearance through the liver and kidneys at 2 and 6 h with nearly complete clearance achieved at 24 h postinjection.

Figure 5.

NP1 reduces the spread of neuroinflammation in a CCI mouse model of TBI. (A) T2-weighted fast spin–echo sequences were used to image the spread of secondary injury as visualized by hyperintense regions surrounding the primary damage (arrows). Slices from 1 mm anterior and 1 mm posterior to the impact site are also shown. (B) Quantification of volume of hyperintense secondary damage through all slices revealed faster reduction in NP-treated mice as compared to untreated mice. (C–J) NP1 improves histological outcome in a CCI mouse model of TBI. (C, D) Histological assessment of structure and reactive astrogliosis in (C) CCI and (D) CCI with NP1 treatment at 1 week postinjury. No structural changes in the ipsilateral CA1, CA2, CA3, and dentate gyrus (DG) regions of the hippocampus were observed with H&E staining. GFAP immunostaining revealed a decrease in reactive astrocyte staining in brains from mice treated with NP1 after TBI. Iba1 immunostaining revealed a decrease in activated microglia in brains from mice treated with NP1 after TBI. Scale bars correspond to 0.25 mm. (E) Quantification of GFAP+ cell number per 40× field (0.05 mm²). There was a statistical difference (P = 0.0019) between CCI and CCI + NP as determined by paired t test. (F) Quantification of Iba1+ cell number per 40× field (0.05 mm²). There was a statistical difference (P = 0.0638) between CCI and CCI + NP as determined by paired t test. (G, H) Histological assessment of structure and reactive astrogliosis in (G) CCI and (H) CCI with NP1 treatment at 1 month postinjury. GFAP and Iba1 immunostaining revealed a decrease in reactive astrogliosis staining in brains from mice treated with NP1 after TBI. Scale bars correspond to 0.25 mm. (I) Quantification of GFAP+ cell number per 40× field (0.05 mm²). There was a statistical difference (P = 0.0481) between CCI and CCI + NP as determined by paired t test. (J) Quantification of Iba1+ cell number per 40× field (0.05 mm²). There was a statistical difference (P = 0.0133) between CCI and CCI + NP as determined by paired t test.

Figure 6.

NP1 improves functional outcome in a CCI mouse model of TBI. (A) Animal weights over 3 weeks from before and after the day of CCI surgery (day 0). A significant drop in weight was observed for animals receiving the CCI surgery at days 1 and 4, but animal recovery was improved in animals that received NP1 treatment. (B) Rotor-Rod analysis of motor function. Mice were trained on the Rotor-Rod (0–50 rpm over 5 min) for 1 week prior to CCI surgery. Mice that received CCI surgery or CCI surgery plus NP treatment showed a significant reduction in latency to fall at 1 day following the surgery as compared to control animals (adjusted P = 0.0428 and 0.0543, respectively, as determined by one-way ANOVA with Tukey’s multiple comparisons test), which fully recovered by day 4, with CCI and CCI + NP mice showing similar performance. (C) Startle response habituation test 1 week after TBI. Both control and CCI + NP mice showed nonzero slopes through regression analyses (P values of 0.0143 and 0.0713, respectively), whereas CCI mice did not show a nonzero slope (P = 0.9851). (D) There were no significant changes in startle amplitude, indicating no change in the reflex circuitry in these mice (P = 0.2569 by one-way ANOVA). (E) Histological analyses confirm there were no histopathological markers of astrogliosis in the PnC region, which is involved in the startle response. (F) Startle response habituation test 1 month after TBI. CCI + NP mice showed a nonzero slope through regression analysis (P = 0.0704), whereas CCI mice did not show a nonzero slope (P = 0.9851). (G) There was a significant difference in startle amplitude at 1 month post-TBI (P = 0.0402 by one-way ANOVA) with CCI significantly lower than CCI + NP (adjusted P value = 0.0370 by Tukey’s multiple comparisons test). Additionally, CCI mice showed significantly lower startle amplitude at 1 month post-TBI as compared to 1 week post-TBI (P = 0.0012 by unpaired t test), whereas startle amplitudes for control and CCI + NP mice were similar between these two time points (P = 0.8699 and 0.9420, respectively, by unpaired t test). (H) Histological analyses showing histopathological markers of astrogliosis in the PnC at 1 month post-TBI.

Scheme 1.

Synthesis of PS80-Based Core-Cross-Linked NPs via Sequential Thiol–Ene and Thiol–Michael Reactions