Article www.acsnano.org Graphene Oxide Nanosheets Reduce Astrocyte Reactivity to Inflammation and Ameliorate Experimental Autoimmune Encephalomyelitis Giuseppe Di Mauro,# Roberta Amoriello,# Neus Lozano, Alberto Carnasciali, Daniele Guasti, Maurizio Becucci, Giada Cellot,* Kostas Kostarelos, Clara Ballerini,* and Laura Ballerini* See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles. Cite This: ACS Nano 2023, 17, 1965−1978 Read Online Downloaded via 139.47.116.106 on October 31, 2023 at 16:41:33 (UTC). ACCESS Metrics & More Article Recommendations * sı Supporting Information ABSTRACT: In neuroinflammation, astrocytes play multi- faceted roles that regulate the neuronal environment. Astrocytes sense and respond to pro-inflammatory cytokines (CKs) and, by a repertoire of intracellular Ca2+ signaling, contribute to disease progression. Therapeutic approaches wish to reduce the overactivation in Ca2+ signaling in inflammatory- reactive astrocytes to restore dysregulated cellular changes. Cell-targeting therapeutics might take advantage by the use of nanomaterial-multifunctional platforms such as graphene oxide (GO). GO biomedical applications in the nervous system involve therapeutic delivery and sensing, and GO flakes were shown to enable interfacing of neuronal and glial membrane dynamics. We exploit organotypic spinal cord cultures and optical imaging to explore Ca2+ changes in astrocytes, and we report, when spinal tissue is exposed to CKs, neuroinflammatory-associated modulation of resident glia. We show the efficacy of GO to revert these dynamic changes in astrocytic reactivity to CKs, and we translate this potential in an animal model of immune-mediated neuroinflammatory disease. KEYWORDS: hemichannels, reactive astrocytes, graphene oxide, drug delivery, nanomedicine N egative and maladaptive features of neuroinflamma- core mediators common to a vast range of CNS patholo- tion contribute to the initiation and development of gies.1,10 diverse diseases of the central nervous system Graphene-based nanotechnology approaches are increas- (CNS),1 which include chronic neurodegenerative pathologies ingly attracting the attention of biomedical scientists11,12 and such as multiple sclerosis, Alzheimer’s and Parkinson’s can offer innovative solutions in neurology.13,14 In this context, graphene oxide (GO) is emerging as a potential multifunc- diseases,2−4 or acute injuries such as stroke and traumatic tional platform for therapeutic neuromodulation.15−17 We have injury.5,6 CNS inflammation is fueled by the production of previously reported the ability of thin, medical grade, and mediators, ultimately released by CNS resident glia (microglia endotoxin-free GO nanosheets with small (100−400 nm) and astrocytes), endothelial cells, and peripherally derived lateral dimensions (s-GO) to selectively target CNS subcellular immune cells. Astrocytes, the key regulators of CNS homeo- components, such as synapses18,19 or modulate astrocyte stasis,7 are active players in neuroinflammation, and recent research has shown that these CNS cells, by connexin Received: July 5, 2022 hemichannel (HC) increased permeability, release pro- Accepted: January 10, 2023 inflammatory factors and may contribute to CNS damage Published: January 24, 2023 progression.8,9 Gaining insight into the cellular mechanisms that sustain neuroinflammation spreading is instrumental to development of future therapeutic tools, in particular to target © 2023 The Authors. Published by American Chemical Society https://doi.org/10.1021/acsnano.2c06609 1965 ACS Nano 2023, 17, 1965−1978 ACS Nano www.acsnano.org Article Figure 1. s-GO characterization and their impact on the morphology of spinal neuroglia exposed to pro-inflammatory CKs. Characterization of s-GO nanosheets. (A) Height AFM image. (B) SEM micrograph. (C) Summary of physicochemical properties by different analytical, spectroscopic, and microscopic techniques. (D) Representative confocal micrographs of ventral areas in spinal cultures with astrocytes and microglia visualized using anti-GFAP (in green) and anti-Iba1 (in red) antibodies. Bar plots summarize (E) GFAP fluorescence intensity, (F) Iba1 positive cell density, and (G) Ti; *** P < 0.001. membrane dynamics.20 More recently, we exploited the anxiety disorders in rats.21 Despite such promising advances in potential application of s-GO by targeting in vivo dysfunctional s-GO control over synaptic activity, relatively fewer studies brain excitatory synapses and preventing the development of investigated the interaction of graphene-based materials, in 1966 https://doi.org/10.1021/acsnano.2c06609 ACS Nano 2023, 17, 1965−1978 ACS Nano www.acsnano.org Article Figure 2. sGO prevent the emergence of CKs-induced dysfunctional Ca2+ signaling in astrocytes and altered synaptic signaling in neurons. (A) Representative fluorescent traces depicting astrocytes calcium activity in control, sGO, CKs, and CKs+sGO conditions. (B) Snapshot of the ventral area visualized by Fluo-4 AM, ROIs highlight active cells (control culture). (C) Bar plot of active astrocytes, (D) box-plot of calcium events frequency, and (E) cumulative distribution of their inter-event intervals. (F) Bright field image of a ventral interneuron targeted by a patch pipette. (G) Representative voltage clamp tracings in control, sGO, CKs, and CKs+sGO conditions, and cumulative probability plots of PSCs (H) inter-event intervals and (I) amplitudes. * P < 0.033, ** P < 0.002, *** P < 0.001. particular s-GO, with CNS glial cells,22 and it has not been experimental autoimmune encephalitis (EAE30,31), one of the addressed whether s-GO could interfere with astrocyte most reliable rodent models for multiple sclerosis where, upon reactivity during neuroinflammation. repeated intravenous delivery, we report that s-GO preserved Here, we explore, by live imaging astrocyte Ca2+ signal its ability to reduce spinal astrogliosis and control neuronal cell dynamics in spinal-slice cultures, an in vitro system where the loss in vivo. More interestingly, such histological changes were sensory-motor cytoarchitecture is preserved in 3D tissue associated with an overall improved clinical score. retaining spinal cord resident cells.23,24 Incubation (6 h) with a pro-inflammatory cytokines (CKs) cocktail (TNF-α, IL-1β RESULTS AND DISCUSSION and GM-CSF25−29) ignited aberrant Ca2+ signaling in astrocytes, accompanied by morphological changes in reactive Graphene Oxide Synthesis and Characterization. The resident glia, altered synaptic activity, and increased connexin s-GO material used in these studies has been synthesized HC permeability. These functional markers of inflamma- according to the protocols that have already been reported and tion25,26 were all effectively down-regulated by s-GO (6 h) characterized extensively in our previous published incubation, and s-GO specifically impaired HC permeability work.18,19,32,33 The specific batch of GO nanosheets used when pharmacologically increased. Last, we attempted to here was thin (1−3 layers) characterized by lateral dimensions translate this protective effect of s-GO in a mouse model of of 100−400 nm, with reproducibly purity and well 1967 https://doi.org/10.1021/acsnano.2c06609 ACS Nano 2023, 17, 1965−1978 ACS Nano www.acsnano.org Article Figure 3. s-GO modulate spinal astrocytes HCs activation by pro-inflammatory CKs. (A) Superimposed fluorescent traces of Ca2+ oscillations in synchronized or unsynchronized pairs of astrocytes recorded in control, s-GO, CKs, and CKs+s-GO conditions. (B) Snapshot of astrocytes pair in Fluo-4 AM. Bar plots summarize the (C) mean astrocyte pair distance and (D) CI. (E) Confocal micrographs of spinal cultures labeled by LY and GFAP. Bar plots summarize (F) LY-GFAP colocalization and (G) LY fluorescence intensity. * P < 0.033, ** P < 0.002, *** P < 0.001. demonstrated biotolerability. Figure 1A−C describe in detail 10 ng/mL) reported to be released in EAE and to induce CNS the physicochemical characterization of the batch of GO reactivity.25−29 We (co)incubated separate groups of untreated nanoflakes used in these studies. or CK-treated cultures with s-GO (6 h) at a concentration of Our previous results in vitro and in vivo strongly support the 25 μg/mL.18,19,21,33 In all four culture groups we used the use of s-GO with this specific range of dimensions in complex specific markers, glial fibrillary acidic protein (GFAP) and biological environments,18,19,21,33,34 and in particular due to s- ionized calcium-binding adapter molecule 1 (Iba1), to monitor GO lack of in vitro and in vivo cytotoxic responses.33,35−37 by immunofluorescence microscopy the morphology of Glial Cell Reactivity Due to Inflammatory Danger Is astrocytes and microglia, respectively (Figure 1D). Controlled by Graphene Oxide. We used organotypic CKs induced a massive enhancement in GFAP mean spinal cord cultures to monitor resident astrocytes and fluorescence intensity (MFI; from 315 ± 27 au in control to microglia reactivity to neuroinflammation and to explore s- 545 ± 30 au in CKs; n = 9 fields each condition; Figure 1D,E) GO impact on activated neuroglia. Spinal explant cultures are a and significantly increased Iba1-microglia cell density (11 ± 1 convenient in vitro model to reproduce an inflammatory cells/field control and 26 ± 1 cells/field CKs, P < 0.001; environment and to gain insight into CNS responses to danger Figure 1D,F) with a typical reduction in Iba1-positive cell signals.25−27 To activate resident glial cells, we adopted a well- transformation index (Ti; 7.09 ± 0.61 control and 1.81 ± 0.08 characterized neuroinflammation paradigm where organotypic CKs, P < 0.001; Figure 1G), supportive of CKs-induced active cultures (2 weeks in vitro, WIV) were incubated (6 h) with a (ameboid) microglia.38−40 All these changes are in accordance mixture of pro-inflammatory CKs (TNF-α, IL-1β, GM-CSF, with astrocyte and microglia activation upon inflammation in 1968 https://doi.org/10.1021/acsnano.2c06609 ACS Nano 2023, 17, 1965−1978 ACS Nano www.acsnano.org Article Figure 4. s-GO control of CKs induced or QN induced Ca2+ events increase in spinal cord astrocytes. (A) Schematic representation of the experimental design. (B) Calcium tracings in control, CKs, and CKs+sGO astrocytes. (C) Bar plot of number of active cells and (D) box plot of calcium events frequency. (E) Schematic representation of the experimental design. (F) Representative fluorescent traces depicting astrocytes calcium events in all conditions. (G) Bar plot of number of active cells and (H) box plot of calcium events frequency. * P < 0.033, ** P < 0.002, *** P < 0.001. spinal explant cultures.25,26 s-GO incubation per se did not to respond to a variety of external stimuli.41,42 Slow calcium alter glia morphology in terms of GFAP MFI (383 ± 25 au in n oscillations, generated by astrocytes in the spinal cord,43,44 = 9), Iba1-positive cell density (13 ± 1 cells/field), and Ti were isolated and measured in the ventral horn of organotypic (5.45 ± 0.41, P > 0.033 control vs s-GO; Figure 1C,E,F). cultures (2 WIV) during calcium imaging experiments with the When coapplied to CKs, s-GO prevented the increase in MFI calcium dye Fluo-4 AM, and in the continuous presence of of GFAP (356 ± 14 au, n = 9 fields) as well as Iba1-positive tetrodotoxin (TTX, 1 μM; see Experimental Section) a sodium microglia changes in cell density (12 ± 1 cells/field) and Ti fast-inactivating blocker that abolished action potentials and (5.38 ± 0.62; Figure 1C−F; P < 0.001 CKs vs CKs+s-GO). removed back-ground neuronal activity.25,45Figure 2A shows These results suggested that s-GO nanosheets exhibited a fluorescent tracings depicting intracellular calcium dynamics in modulatory action on the development of neuroinflammation astrocytes recorded from the sampled area (320 × 320 μm2, in spinal cord slices. Figure 2B) of the ventral spinal horn in control and CKs s-GO Protect Spinal Astrocytes and Neurons from cultures, with or without s-GO. In control glial population, Dysfunctional Signaling Triggered by CKs. Astrocytes, via calcium dynamics emerge as slow oscillations occurring at a complex spatial-temporal changes in cytoplasmic Ca 2+ low pace. When in the presence of s-GO, the number of active concentrations, integrate and propagate signals in the CNS cells and the frequency of calcium oscillations did not change 1969 https://doi.org/10.1021/acsnano.2c06609 ACS Nano 2023, 17, 1965−1978 ACS Nano www.acsnano.org Article (Figure 2C−E). Resident glial cell reactivity to pro- s-GO coapplication (0.64 ± 0.07, n = 6; CKs vs CKs+s-GO P inflammatory CKs more than doubled the amounts of active < 0.002). astrocytes (from 9 ± 1 cells/field in n = 12 control to 24 ± 2 We used the low molecular-weight dye lucifer yellow (LY; 1 cells/field in n = 10 CKs, P < 0.001, Figure 2B,C) and strongly mM) to quantify HC permeability47,48 in astrocytes in all four increased events frequency (from 0.008 ± 0.001 Hz in control conditions. In Figure 3E, representative micrographs show LY to 0.034 ± 0.003 Hz in CKs, P < 0.001, Figure 2A−D). s-GO, and GFAP colabeling, summarized in the plot in Figure 3F. CK when coapplied to CKs, efficiently prevented the boost in treatment resulted in a statistically significant (P < 0.001) astrocyte calcium oscillations, leaving a basal activity increase in double LY-GFAP positive cells when compared to reminiscent of control (11 ± 1 cells/field and 0.011 ± 0.001 control (11 ± 1 cells/field control and 31 ± 1 cells/field CKs; Hz, n = 8, P < 0.001; Figure 2A−D). In Figure 2E, the changes n = 9 fields each; Figure 3E,F). Such an increase was prevented in calcium dynamics are summarized as inter-event intervals in CKs+s-GO (12 ± 1 cells/field CKs+s-GO; CKs vs CKs+s- and compared among the culture groups as cumulative GO, P < 0.001), while s-GO per se did not differ from control probability plots (control, s-GO and CKs+s-GO vs CKs P < (10 ± 1 cells/field s-GO, n = 9 fields each Figure 3E,F). 0.002; s-GO, and CKs+s-GO vs control P > 0.033). Similarly, s-GO coapplied with CKs prevented the increase in In organotypic cultures, astrocyte aberrant calcium signals LY fluorescence intensity summarized in Figure 3G (MFI 34 ± contribute to cytokines and chemokines release due to 7 au control, 108 ± 6 au CKs and 38 ± 4 au CKs+s-GO; CKs inflammation.25 CKs directly or via resident cells boost vs control and vs CKs+s-GO P < 0.001). network excitability resulting in an enhanced occurrence of When preincubated in carbenoxolone (CBX, 200 μM, 10 spontaneous postsynaptic currents (sPSC26,27). min), an antagonist of gap-junction (GJ)49,50 and of active Patch clamped ventral interneurons (Figure 2F) displayed HCs,51 we detected a reduced dye uptake in all conditions, an intense basal synaptic activity in control,24,46 and with no subsequent increase by CKs (Figure S2A,B). These representative tracings in Figure 2G show the occurrence of results further supported that s-GO nanosheets modulated heterogeneous sPSCs, which were not affected by s-GO (sPSC astrocyte reactivity to the inflammatory milieu in spinal slices, inter event intervals from 188 ± 26 ms in controls n = 9 to 236 suggesting their specific ability to block HC increase in ± 36 ms in s-GO n = 8; sPCS amplitude from 19 ± 1pA in permeability. control to 21 ± 1pA in s-GO Figure 2G−I). Conversely, CKs Quinine Activation of Connexin HC Induces Aberrant treatment induced a strong increase in synaptic activity, Calcium Signals Blocked by s-GO. In the next set of completely prevented by s-GO coincubation (sPSC inter event experiments, s-GO was added after 2 h of inflammation induction by CK treatment (Figure 4A27) to challenge its intervals from 55 ± 5 ms in CKs n = 8 to 166 ± 22 ms in CKs ability to control aberrant Ca2+ signals and to exclude any +s-GO n = 7; sPCS amplitude from 27 ± 1 pA in CKs to 21 ± direct interference of s-GO with CKs treatment such as 1 pA in CKs+s-GO; cumulative distribution in Figure 2H, adsorption on flakes surfaces. In Figure 4B, fluorescence control, s-GO and CKS+s-GO vs CKs < 0.001, in Figure 2I tracings depict glial responses to CKs, in terms of increased control, s-GO vs CKs P < 0.001, CKs vs CKs+s-GO P < number of active cells and oscillation frequency (bar plot and 0.033). box-plot in Figure 4C and D, respectively). The delayed The protective effects of s-GO on CKs-mediated over- application of s-GO could still efficiently decrease the number activation of calcium signaling in glial cells, and of synaptic of CK active glial cells (13 ± 1 cells/field control, 25 ± 4 cells/ signaling in neurons, strengthen the hypothesis of s-GO field CKs and 9 ± 1 cells/field CKs + s-GO treated ones, n = 8 modulatory action over neuroinflammation. Such a modulatory each, P < 0.001; Figure 4C) and was able to tune back to role is also supported by s-GO ability to control the production control levels the oscillation frequency (0.016 ± 0.002 Hz and release of cytokines (IL-1β; IL-6; TNF-α) and chemokine control, 0.054 ± 0.004 Hz CKs and 0.017 ± 0.001 Hz CKs+s- (CCL2; CXCL10) measured by Milliplex assay of the GO, P < 0.001, Figure 4B−D). supernatants harvested from organotypic slices25 in CKs or We used quinine (QN; Figure 4E) to directly activate CK+s-GO treated slices (Figure S1). connexin HC.47Figure 4F−H show that QN (200 μM, 4 h) s-GO Treatment Prevents CKs-Induced Increase in increased the number of active glial cells from 10 ± 1 cells/ Glial Hemichannels Opening. In GFAP-positive cells, pro- field (n = 5 control) to 20 ± 2 cells/field in QN (n = 5; P < inflammatory CKs are reported to increase the permeability of 0.002, Figure 4G), as well as calcium oscillation frequency membrane hemichannels (HC), a pathway for diffusion of (from 0.014 ± 0.002 Hz in control to 0.040 ± 0.005 Hz in small molecules or ions contributing, although not exclusively, QN, P < 0.002; Figure 4F−H), mimicking aberrant calcium to inflammation spread and to aberrant asynchronous calcium signaling brought about by CKs. QN-induced dysfunctional signaling among astrocytes in cultured spinal explants.25 We calcium signals were not detected when s-GO was coapplied quantified Ca2+ oscillations synchrony by measuring the cross- after 2 h with QN (11 ± 1 cells/field and 0.019 ± 0.002 Hz correlation index (CI; see Experimental Section) between pairs QN+s-GO, n = 5; significance vs QN P < 0.033 and P < 0.002, of active astrocytes. In Figure 3A, superimposed fluorescence respectively; Figure 4F−H). Thus, QN calcium dysregulation tracings are typical examples of simultaneously recorded pairs due to the opening of connexin HCs was blocked by s-GO, of glial cells (visualized by Fluo-4; Figure 3B) located at suggesting a direct impact on HC permeability. Such a comparable distance in the four conditions (bar plot in Figure hypothesis is further strengthened by the ability of s-GO to 3C). As summarized by the bar plot in Figure 3D, control and prevent the increase in double LY-GFAP positive cells s-GO treated cells showed synchronized oscillations with following QN application (Figure S3A−C). similar CI values of 0.49 ± 0.06 in control (n = 8) and 0.69 ± s-GO Treated EAE Mice Show Milder Disease, 0.11 in s-GO (n = 6; P > 0.033). In CKs, glial cells displayed a Reduced Astrogliosis, and Increased Neuronal Survival. statistically significant (P < 0.033) lower CI (0.22 ± 0.02 in The capacity of s-GO nanosheets to modulate neuro- CKs, n = 8); such a desynchronization was fully prevented by inflammation in vivo was tested in EAE mice, which upon 1970 https://doi.org/10.1021/acsnano.2c06609 ACS Nano 2023, 17, 1965−1978 ACS Nano www.acsnano.org Article Figure 5. s-GO treatment in EAE mice ameliorates disease progression, reduces astrogliosis, and protects neuronal survival in the spinal gray matter. (A) At 7 days postimmunization (d.p.i.), EAE mice were treated by intravenous (i.v.) tail vein injections with 100 mL of PBS (control) or with s-GO (sGO) at 7, 10, 13, and 16 d.p.i. and sacrificed at 22 d.p.i. (B, C) Clinical score and weight (daily, 9−22 d.p.i.) of control and sGO EAE mice. (D) Fluorescence micrographs of spinal cord sections labeled with DAPI (in blue) and GFAP (in green). (E) Bar plot of the percentage of GFAP+ cells in control and in sGO EAE spinal gray matter. (F) Low magnification of EAE spinal sections in control and sGO EAE mice labeled by GFAP (in green). (G) Fluorescence micrographs of spinal cord sections labeled with DAPI (in blue) and Iba1 (in red). (H) Bar plot summarizes microglia Ti values in control and in sGO EAE spinal gray matter. (I) Fluorescence micrographs of spinal cord sections labeled with DAPI (in blue) and NeuN (in red). (J) Bar plot summarizes the percentage of NeuN+ cells in control and in sGO EAE spinal gray matter. (K) TEM micrographs of EAE mice spinal cord sections in (i, ii) control and (iii, iv) sGO; (v) box area: Raman spectrum in EAE sGO mice. * P < 0.033, ** P < 0.002, *** P < 0.001. 1971 https://doi.org/10.1021/acsnano.2c06609 ACS Nano 2023, 17, 1965−1978 ACS Nano www.acsnano.org Article MOG35−55 immunization developed neuroinflammation-driven while similar aggregates were never detected in CTRL EAE clinical and histological alterations, hallmarks of EAE (see (Figure 5K, panels i and ii). Raman spectroscopy of such Experimental Section31). After MOG35−55 administration, mice deposits confirmed a profile compatible with that of s-GO were treated with s-GO (4 mg/kg in 100 μL PBS; n = 8 (Figure 5K, panel v). In agreement to that reported when animals) or PBS (100 μL, CTRL; n = 8 animals) through multiple sclerosis animal models are treated with nanomaterials multiple tail vein injections at 7, 10, 13, and 16 days of comparable dimensions,54 we suggest that in EAE mice the postimmunization (d.p.i., Figure 5A). compromised blood−brain barrier (BBB)55,56 allows s-GO EAE onset and progression were monitored by daily nanosheets to pass the BBB and target the CNS, reducing evaluation of the clinical score (see Experimental Section) astrogliosis and neuronal death in the spinal cord, without and animal weight for 3 weeks. Despite the comparable clinical affecting the peripheral immune system (Figure S4). values at 9 d.p.i. in CTRL and s-GO EAE mice (0.18 ± 0.13 Our evidence suggests that s-GO block the inflammation- CTRL and 0.06 ± 0.06 s-GO; P > 0.033; Figure 5B), in the induced calcium dynamics in in vitro astrocytes mainly by following days CTRL animals showed a strong increase interfering with the activation of connexin HC.25 In addition, (worsening) of the clinical score, in agreement with the we demonstrate that the s-GO protective effect could be progressive development of the disease,52 while s-GO translated from cultured spinal explants to an in vivo model of presented a significantly less severe progression (at 14 d.p.i. EAE inflammatory disease. 1.8 ± 0.18 CTRL and 1.1 ± 0.2 s-GO, P < 0.033; at 22 d.p.i. 2 The exposure of spinal explants to a CKs cocktail that ± 0.1 CTRL and 1 ± 0.1 s-GO, P < 0.002; Figure 5B). The mimics pro-inflammatory conditions associated with neuro- significantly lighter EAE disease progression in s-GO-treated degenerative disorders57,58 activates neuroinflammation onset animals was accompanied by a trend of more stable body and propagation, as shown by GFAP-positive astrocytes, Iba1- weight, although not statistically significant when compared to positive microglia, astrocyte-calcium, and synaptic network CTRL progressive loss (Figure 5C). reactivity.25−27 These pathological aspects involving different To investigate the histological hallmarks of EAE, animals cell types are reverted by s-GO treatment, indicating that the were sacrificed at 22 d.p.i. and we collected for analysis the nanomaterial could target an upstream key component of the spinal cord, the draining lymph nodes, and the spleen. In the pro-inflammatory cascade. Reactive astrocytes in cultured spinal cord, we measured astrogliosis, microglial activation, and slices are identified by the functional features of calcium neuronal viability. In s-GO EAE spinal cord, gray matter GFAP dynamics (resistant to TTX, low pace kinetics, enhanced by positive cells were decreased (in %) when compared to CTRL CKs;25 or by HC activators47) and cell location (close to EAE (17 ± 1% CTRL and 10 ± 1% s-GO, P < 0.001; Figure ventral neurons25). We focus on aberrant calcium activity, 5D,E), while no differences were found in white matter which mediates inflammatory cytokines and chemokines astrogliosis (Figure 5F; measured as corrected total cell release in response to danger signals,25 as a hallmark of fluorescence, CTCF, that was 4654074 ± 241754 au CTRL reactive astrocytes, and we suggest this as the main target of s- and 4081208 ± 243423 au s-GO; P > 0.033). Iba1-positive GO protective effects. microglia Ti did not differ in the two EAE groups, with values s-GO flakes were reported to interface exoendocytotic indicative of ameboid reactive cells (Ti: 12 ± 1 CTRL and 11 membrane dynamics at the nanoscale, interfering with ± 1 s-GO, P > 0.033; Figure 5G,H). shedding of membrane vesicles in astrocytes20 and with Neuronal viability, classically affected by neuroinflammation presynaptic vesicle release in neurons.18 These effects, damages driven by EAE,53 was measured as % of viable potentially involved in tuning spinal circuit inflammatory neurons, identified by the specific neuronal nuclei marker reactivity, are unlikely in the current study, since both require NeuN. Figure 5I and J show an increase in NeuN+ neurons in prolonged exposure (days to weeks) to localized delivery of s- s-GO treated vs CTRL (17 ± 1% CTRL and 26 ± 1% s-GO GO at high concentrations18,20,59 and, in spinal explants, injected animals; P < 0.001). We additionally reported lymph involve the activation of microglia.59 Accordingly, in the node cell viability and phenotype as percentages of T, B, and current tests, in the absence of neuroinflammation, s-GO per dendritic cells (CD3+, CD19+, and CD11c+ cells, respec- se do not modulate neuroglia morphology, astrocyte calcium tively). In comparing CTRL with s-GO EAE animals, we did dynamics, or synaptic activity. We also exclude a direct not detect statistically significant differences in lymph node cell interference, such as a mere adsorption of pro-inflammatory death (% propidium iodide, PI, positive cells: 11 ± 1% CTRL cues by s-GO flakes, since the nanomaterials are effective even and 9 ± 1% s-GO, P > 0.033; Figure S4A) and in the when added 2 h after CKs exposure. phenotype of viable cells (by flow cytometry, for CD3+ cells We can speculate that s-GO, which feature adhesion to 33 ± 2% CTRL and 28 ± 2% s-GO; for CD19+ cells 35 ± 3% complex patches of cellular membranes,60 specifically target CTRL and 34 ± 3% s-GO; for CD11c+ cells 1 ± 0.2% CTRL HC permeability in reactive cells. and 0.6 ± 0.1% for s-GO, all the P-values were > 0.033; Figure Nonjunctional HCs in astrocytes activate upon inflamma- S4B). Figure S4C shows the proliferation responses of tion,61,62 allowing the pathological flow of ions and molecules, splenocytes and lymph node cells upon stimulation with which promote neuroinflammation spreading.8,63 In our MOG35−55 or with the specific mitogen phytohemagglutinin previous study, we reported that increased HC permeability (PHA). In all cases, the exposure to s-GO did not impair the was feeding aberrant calcium signaling and contributed to peripheral immune system cell viability or proportions between astrocyte inflammatory status in spinal explant slices.25 Here, different phenotypes, and stimulated s-GO immune cells are pro-inflammatory HC activation is supported by CKs even more prone to proliferate. recruiting of active astrocytes with enhanced and desynchron- To explore whether s-GO reached the spinal cord in s-GO ized calcium oscillations25 together with the increased LY EAE, we used transmission electron microscopy (TEM; see uptake in GFAP positive cells, an assay used to monitor HC Experimental Section). Such analysis revealed material deposits permeability.47,48 LY uptake is significantly reduced by CBX, a in s-GO EAE mice spinal cords (Figure 5K, panels iii and iv), HC inhibitor,51 strengthening the involvement of such proteins 1972 https://doi.org/10.1021/acsnano.2c06609 ACS Nano 2023, 17, 1965−1978 ACS Nano www.acsnano.org Article in reactive astrocytes calcium signaling pathways. Indeed, in vivo controlling neuroinflammation neuronal damage might the absence of the inflammatory milieu, the application of QN, impact future strategies to treat CNS pathological conditions. a HC activator,47 induces changes in astrocyte calcium dynamics reminiscent of those provoked by neuroinflamma- LIMITATIONS OF THE STUDY tion, efficiently blocked by s-GO. Although we cannot exclude Multiple sclerosis is a human complex, multifocal disease that the contribution of other membrane pathways to calcium features distinct disabilities. Astrocytes regulate neuroinflam- microdomains, our results strengthen the link among HCs matory responses implicated in the development of EAE opening, calcium signals in reactive astrocytes, and inflamma- disease, an animal model of multiple sclerosis. In EAE, through tion,64 and we propose that s-GO protective effects can be active gliosis and upregulation of immune related genes72,73 mediated by a reduction in HC permeability, interrupting a astrocytes, in particular those located in the spinal cord, vicious cycle in the spreading of neuroinflammation. We assess contribute to disease progression and may represent potential s-GO treatment of EAE mice, modeling multiple sclerosis, a targets for therapeutic intervention.1,10 In our study, we well-known neuroinflammatory disease.63 In EAE mice, the characterize in vitro astrocyte reactivity when involved in disorder arose and progressed within few weeks, showing well- neuroinflammation progression ignited by transient CKs characterized patterns of histological and clinical aspects65 exposure. Since we measured cell calcium responses and we including astrogliosis and neuronal loss in the spinal cord-gray monitored the emergence of gliosis in spinal cord explants, we matter.65 s-GO delivery, without altering the function of the propose a mechanism through which s-GO, by targeting peripheral immune system, reached the spinal tissue and specialized membrane channels, prevent dysregulated calcium control astrocyte reactivity in the spinal cord gray matter, signaling and the neuroinflammation cascade involving favoring neuronal survival. Such tissue protection is accom- astrocytes. We further test s-GO in EAE mouse in vivo, in panied by an ameliorated clinical phenotype. To note, in EAE, which repeated administrations of s-GO reduced astrogliosis due to the repeated s-GO administrations, we cannot rule out and neuronal cell loss in the spinal cord, accompanied by s-GO control over CNS synaptic activity,18,21 contributing to ameliorated clinical signs. spinal neurons protection against neuroinflammation.26 Despite these encouraging results, our in vivo observations Nanomaterials are increasingly engineered to treat neuro- remain limited to a proof-of-concept experiment which might pathologies including neuroinflammation.66 In preclinical indicate novel targets, channel permeability in astrocytes, and animal models, such as EAE, such engineered nanoparticles unconventional tools, nanomaterials, to optimize neuro- were developed for targeted delivery of tolerogenic molecules protective treatment developments. Translational implications to reduce human autoimmunity in an antigen-specific of s-GO are still in their infancy due to the lack of in-depth manner67 or more in general to favor therapeutic drug delivery pharmacological investigations or of ad hoc studies addressing based on nanotechnology approaches.68 For example, lip- potential toxicity when using s-GO based vectorized systems in osomes, polymeric nanoparticles, dendrimers, and dispersed EAE. Dedicated pharmacokinetic and pharmacodynamic carbon nanotubes were engineered to vectorize anti-inflam- studies are warranted. Furthermore, the current study should matory drugs and improve their pharmacokinetic properties; be extended to other animal species and explore if there is any unfortunately, cell toxicity issues emerged when using many of sex dependent effect. these engineered nanomaterials.69 Differently, s-GO showed an In general, the biological tolerability and biocompatibility of intrinsic ability to modulate reactive cells, in the absence of cell s-GO injectable suspensions has been addressed in numerous toxicity, as also shown by γ-H2AX assay70 (Figure S5), ̈ studies performed previously using the same material in naive, excluding nuclear damage. healthy animals. The rapid excretion, upon intravenous Although promising, currently the clinical translation of administration, through urine of a large fraction of the s-GO nanomaterial-based therapeutic approaches for the treatments nanosheets74 and the absence of pathogenic alterations in the of nervous system pathologies is limited by the lack of standard organs of accumulation,75 are suggestive of an overall biological procedures for assessing the toxicology of these novel profile for s-GO that consists of elimination and degradation compounds.71 How long s-GO persist in the body and their by endogenous enzymes76,77 with no consequent adverse potential translocation to secondary vital organs or nodules of effects. the lymphatic system are all factors that should be considered in future clinical studies. EXPERIMENTAL SECTION CONCLUSIONS Graphene Oxide Nanosheet Preparation. GO was obtained from graphite powder (Sigma-Aldrich, UK) and biology grade Cell-targeting therapeutics in the CNS might take advantage of materials were derived as previously described (Hummers’ nanomaterial-based biointegrated platforms and, more in methods).78 We used non pyrogenic materials to guarantee general, of engineered nanosized materials. In the current endotoxin-free GO as previously described.32,79 work, we exploit GO nanoflakes with small lateral size (s-GO) Atomic force microscopy. Atomic force microscope (Asylum to target the over-reactivity of astrocyte, an early event linked MFP-3D, Oxford instruments), standard air-tapping mode (silicon to neuroinflammation spreading, and a shared hallmark of probes Ted Pella, silicon probes), 300 kHZ resonance frequency, and CNS disorders. s-GO flakes lessen astrocyte reactivity to 40 N/m force, was used. We used mica surface (Ted Pella, freshly inflammation by downregulating hemichannel opening and cleaved) layered by 20 μL of poly-L-lysine 0.01% solution (Sigma- aberrant calcium signaling. We translated our in vitro findings Aldrich), and, upon wash with water, 20 μL of GO suspension (100 μg/mL) was drop casted. Gwyddion software (http://gwyddion.net, to an in vivo mouse model of immune-mediated neuro- version 2.56) was used to image processing. ImageJ software (https:// inflammation disease, where s-GO reduced astrocyte reactivity imagej.nih.gov) with AFM images 5 μm × 5 μm height distribution and improved neuronal survival. This tissue protection was analysis allowed was used to obtain the lateral dimensions values. accompanied by an ameliorated clinical phenotype. The Scanning Electron Microscopy. Magellan 400 L field emission potential hold by s-GO to target astrocyte dysregulation in scanning electron microscope (Oxford instruments; ICN2 Electron 1973 https://doi.org/10.1021/acsnano.2c06609 ACS Nano 2023, 17, 1965−1978 ACS Nano www.acsnano.org Article Microscopy Unit) was used with Everhart-Thornley (secondary NA), cultures loaded with the Ca2+ sensitive dye were excited with a electrons detector) at 20 kV (acceleration voltage) and 0.1 nA (beam mercury lamp at 488 nm. We used a 395 nm dichroic mirror and DN current). In each sample, 20 μL of GO material (100 μg/mL) was filter (1/32), and images were acquired at 7 fps every 150 ms by an deposited on an Ultrathin C on Lacey C grid. ImageJ software was ORCA-Flash4.0 V2 sCMOS camera (Hamamatsu) and a setup allowed to obtain the lateral dimension distribution. controlled by HCImage Live software. Initial recordings (10 min) in Raman Spectroscopy. Confocal Raman microscope (Witec, at saline solution were made to analyze the spontaneous neuronal and RT) with 632 nm laser excitation and 600 g/nm grating allowed us to glial activity prior to tetrodotoxin (TTX, 1 μM, Latoxan;25) acquire Raman spectra. Single Raman spectra were collected (from application. Under these recording conditions, astrocyte activity was several spots, 1 mW and 10 s irradiation) using a 50× objective. Drop visualized25 in the premotor region of the slice ventral zone. The casted samples (20 μL) were analyzed by Origin software. ID/IG recorded images were processed identifying regions of interest (ROIs) intensity ratio was extrapolated by peak intensities with no baseline around cell bodies with Fiji software. Related tracings were transferred correction. to Clampfit software (10.6 version; Molecular Device LLC, US) and Zeta Potential Measurements. Zeta-sizer Nano ZS (Malvern analyzed off-line to quantify the number of oscillating glial cells, the instruments) at the ICN2Molecular Spectroscopy and Optical frequency of oscillations, their IEIs, and the CI. Ca2+ oscillations were Microscopy Facility was used to measure zeta potential (ζ). Each measured as ΔF/F0, where ΔF is the fluorescence increase over the sample (20 μg/mL GO) was measured three times at room baseline, and F0 is the baseline level of fluorescence (as the median of temperature with automatic refractive index and viscosity for water the frame fluorescence values), calculated using the following formula: dispersant settings. Organotypic Spinal Cord Cultures and Treatments. All F F0 100 × experiments were in accordance with the EU guidelines (2010/63/ F0 UE) and Italian law (Decree 26/14) and approved by the authority veterinary service and by our institution (SISSA) animal wellbeing with F, fluorescence value and F0, baseline fluorescence. committee (OBPA). Italian Ministry of Health (no. 22DABNQYA) Luminex Assay and Analysis of Cytokines and Chemokines. animal use approval and EU Recommendation 2007/526/CE. In the supernatants obtained from organotypic slices, a panel of 13 Organotypic spinal cord and dorsal root ganglia slices were CKs and chemokines was analyzed (IFNγ, IL1α, IL1β, IL4, IL6, IL10, obtained as previously described.23,26,27 Experiments of immuno- IL12p40, IL12p70, IL17, C-X-C motif ligand 10 [CXCL10], C-C fluorescence, LY assay, calcium imaging, and patch clamp recordings motif ligand 2 [CCL2], C-X-C motif ligand 2 [CXCL2], TNFα) by were performed after 6 h incubation with the following treatments: (i) means of Milliplex assay (Merck Millipore, USA, #MCYTOMAG- CTRL, standard medium; (ii) s-GO, standard medium and s-GO at 70K-13), and Bio-Plex device (Bio-Rad, USA), as previously 25 μg/mL (54); (iii) CKs, standard medium and a cocktail of the reported.25 CKs and chemokines amounts were expressed as pg/mL mouse recombinant cytokines (10 ng/mL each) TNF-α (R&D with a detection threshold of 1.0 pg/mL for all the analytes. Data were systems, #210-TA/CF), IL-1β (R&D systems, #M15330) and GM- analyzed by calculating differences (Δ) between CTRL and CKs- CSF (R&D systems, #P04141);26,27 (iv) CKs+s-GO, standard stimulated organotypic cultures (ΔCKs − CTRL) or between medium, CKs cocktail, and s-GO at 25 μg/mL. In another set of organotypic cultures stimulated with both CKs and s-GO or only experiments, s-GO were applied 2 h after the application of the CKs with s-GO (Δ [(CKs + s-GO) − s-GO]). Analysis was performed on cocktail (delayed application), for a total time of 4 h coincubation. A three replicates for four organotypic cultures, corresponding to 12 similar time line was used in experiments with quinine (QN, 200 μM, measurements per group. Sigma), where sGO were applied for 2 h after QN application. Electrophysiological Experiments. Electrical activity of inter- The experiments on the genotoxicity of s-GO were carried out neurons, visually identified in the ventral horn80,81 through an exposing slices to s-GO at 25 μg/mL for 6 h. In the case of the inverted microscope (Nikon Eclipse TE200), was recorded by using excitotoxicity damage (positive control), another set of slices was whole cell patch clamp technique. During experiments, slices were treated with glutamate (100 μM) for 1 h, and then fresh medium was perfused with the supra-mentioned standard saline solution. replaced for 24 h before immunofluorescence. Patch pipettes (4−7 MΩ) were filled with (in mM) 120 K Immunofluorescence Imaging and Analysis. Immunofluor- gluconate, 20 KCl, 10 HEPES, 10 EGTA, 2 MgCl2, 2 Na2ATP (pH = escence staining was performed as previously reported.26 Secondary 7.3 with KOH and 295 mOsm of osmolarity). Recordings were antibodies: Alexa 488 goat antimouse (1:500, Invitrogen), Alexa 594 performed at RT. We used a Multiclamp 700A amplifier (Axon goat antirabbit (1:500, Invitrogen). DAPI (Thermo Fisher Scientific) Instruments), 10 kHz sampling, and signals were digitized with the was also added to mark nuclei. We used Fluoromount-G (Invitrogen) pCLAMP software (Molecular Device LLC, US) for off-line analysis. to mount samples and Nikon A1R confocal microscope to acquire In voltage clamp mode, in order to avoid significant distortions of images. In the case of GFAP-Iba1 analysis, we acquired confocal synaptic currents, we considered only recordings with series resistance sections every 1 μm with total Z-stack thickness 5 μm by 40× (0.95 <20 MΩ, not compensated for. Liquid junction potential was −14 NA) objective, while in the case of GFAP/γ-H2AX, we performed mV, and values of potential were not corrected for it. confocal section every 0.5 μm to a total Z-stack thickness of 15 μm by No differences were observed in membrane capacitance (43 ± 4 pF 100× (1.45 NA) objective. Confocal sections were acquired every 1 control, 40 ± 6 pF s-GO, 39 ± 4 pF CKs, and 44 ± 8 pF CKs+s-GO) μm to a total Z-stack thickness of 5 μm. Analysis was performed using and input membrane resistance (425 ± 106 MΩ control, 472 ± 86 Fiji software on three independent culture series. Regarding the MΩ s-GO treated, 263 ± 18 MΩ CKs, and 515 ± 120 MΩ CKs+s- morphology analysis of microglia, we quantified the area and GO) of spinal neurons recorded upon various treatments. perimeter to compute the Ti,26 which is indicative for the microglia Spontaneous activity was monitored at a holding potential of −56 ramification status, while the astrocyte quantification of fluorescence mV. sPSCs analysis was performed using the Clampfit 10.6 software intensity was performed by measuring the mean gray value. The (Molecular Device LLC, US). For each recording, ≥ 100 sPSCs were analysis of H2AX phosphorylation at Ser 139 was carried out by collected, and the inter-event intervals and peak amplitudes were counting the number γ-H2AX foci formation per nuclei colocalizing measured. with GFAP and the number of γ-H2AX/DAPI/GFAP positive cells. Lucifer Yellow (LY) Uptake. For dye uptake experiments, we Live Cell Ca2+ Imaging. Spinal organotypic slices were incubated used LY (Sigma). Cultures were incubated (10 min) at 37 °C with 1 in Fluo-4 AM (4 μM, Invitrogen; 1 h, 37 °C). After the loading and mM of LY in standard physiological solution and washed three times the de-esterification of the dye, the cultures were imaged by an with the same solution without LY. To investigate how GJs and HCs inverted microscope (Nikon Eclipse Ti−U). At room temperature contributed to dye uptake, an independent set of spinal cord slices (RT), they were perfused continuously at 5 mL/min with a saline was previously treated with CBX (Sigma) at 200 μM for 10 min and solution (in mM: 150 NaCl, 4 KCl, 1 MgCl2, 2 CaCl2, 10 HEPES, 10 later incubated with LY. Cultures were fixed with 4% PFA, and glucose (pH = 7.4). Imaged with a 40× objective (PlanFluor, 0.60 subsequently astrocytes were labeled using anti-GFAP as described 1974 https://doi.org/10.1021/acsnano.2c06609 ACS Nano 2023, 17, 1965−1978 ACS Nano www.acsnano.org Article earlier. Images were acquired at the confocal microscope using 20× approximately 0.04 mm2). Astrogliosis in white matter was analyzed and 40× objectives, with a step of 1 μm for a total Z-stack 5 μm by ImageJ software by calculating the CTCF, measured as integrated thickness. Analyses of LY-GFAP double positive astrocytes and the density − (area of a selected cell × mean fluorescence of background quantification of LY fluorescence intensity were performed using Fiji readings). software on three independent culture series. On Iba1-labeled spinal cord sections, microglial activation was Induction of EAE. Experiments were carried out on 6−8 weeks analyzed on selected areas from a range of 5−10 random fields/slide old C57BL/6 female mice (Charles River Laboratories, Italy). using ImageJ software, evaluating the morphology variation of Animals (CeSaL at University of Florence, Italy) were kept at a microglia, to measure the Ti, as previously described in this temperature of 23 °C, 12 h (7am−7 pm) light/dark cycle. Two manuscript. experimental series were performed for a total of 16 mice. All Cytological Analyses: Lymph Node Cells Viability and experiments were authorized according to the Committee for Animal Phenotype. Draining lymph nodes of mice were washed with 1% Care guidelines (d.lgs. 26/2014; authorization n. 1145/2020-PR). of penicillin/streptomycin and 1% fungizone in PBS and mechanically Animals were immunized through three subcutaneous (s.c.) dissociated on sterile Falcon 70 μm cell strainer. Lymph node cells injections (two interscapular and one at the base of tail) of 200 μg were centrifuged 1300 rpm, 10 min, and counted. Cells of lymph of MOG35−55 peptide (MEVGWYRSPFSRVVHLYRNGK, purity of nodes were analyzed by CyFlow Space flow cytometer (Sysmex 85%; Espikem, Italy) solved in sterile water, emulsified with an equal Partec, Gemany) for viability using propidium iodide (PI; 2.5 μg/mL; volume of Complete Freund’s Adjuvant (CFA), and supplemented Molecular Probes, USA) to detect necrotic cells, and for phenotype with 7 mg/mL of Mycobacterium tuberculosis (H37Ra strain; NR by labeling cells for 20 min, at RT, with the following antimouse Nannini, Italy). To increase the BBB permeability, each mouse antibodies: CD3 FITC (clone 145−2C11, BioLegend, USA), CD19 received, the day of immunization and after 48 h, 500 ng (in a volume PE (Miltenyi Biotec, Germany), CD11c PE (clone N418, of 100 μL) of Bordetella pertussis toxin (PTX; Merck, Germany) eBioscience, USA). Flow cytometry data were acquired following dissolved in PBS by intraperitoneal (i.p.) injection. All the standard guidelines82 and analyzed by the FloMax software (Sysmex experiments with animals were blinded. EAE mice were monitored Partec, Germany). CD3+, CD19+, and CD11c+ were gated on a daily for weight and signs of clinical disease, with (0) no signs of forward scatter (FSC) vs a side scatter (SSC) dot plot. Gates were disease; (1) hind limb or limp tail weakness; (2) hind limb and limp labeled according to FloMax software default (Figure S2B; gate “R1” tail weakness; (3) partial paralysis of hind limb; (4) complete paralysis includes the whole reference cell population; gate “Q4” includes of hind limb; or (5) moribund or death by EAE. If reaching the score CD3+, CD19+, or CD11c+ cells within the “R1” gate, whereas “Q3” 4, mice were sacrificed for ethical reasons. includes cells negative for all the above-mentioned markers). In Vivo Treatment with s-GO. Mice were separated in two Statistical Analysis. All the data sets underwent to the normality groups: one treated with PBS as control, and one treated with s-GO. analysis to assess if data were as in a Gaussian distribution. On this The attribution of animals to the experimental groups was basis, concerning the work on organotypic spinal cord slices, the data randomized. s-GO were administered four times (7, 10, 13, and 16 sets were processed with ordinary one-way ANOVA test to measure if days postimmunization, d.p.i.) by intravenous (i.v.) injection in the differences were statistically significant (post hoc test by Holm-Sidak tail vein, at 4 mg/kg of concentration in a PBS volume of 100 mL for comparisons), otherwise Kruskal−Wallis test was used (post hoc test injection. The dose used for the pilot study in vivo was estimated by Dunn’s multiple comparison). Cumulative probability analyses based on our in vitro results (10−25 μg/mL18,21) and on our previous were performed using the Kolmogorov−Smirnov test among groups. in vivo results with local s-GO tissue administration (to the Regarding in vivo experiments, the progression of clinical score and hippocampus 50 μg/mL and to the amygdala 50 μg/mL18,21). In weight were evaluated by using two-way ANOVA test to measure parallel, an equal volume of PBS was administered to control mice. At significant differences among groups, with post hoc test Fisher’s LSD. 22 d.p.i., after mice sacrifice draining lymph nodes, spleen and spinal Statistical significance of all the other results were tested as unpaired t cord were taken for cytological and histopathological analyses. Test (parametric data) or Mann−Whitney (nonparametric data). All Histopathological Evaluation on Mice Spinal Cord. After the data are expressed as Mean ± SEM, n = number of cultures, removal, mice spinal cord was immediately fixed by 4% PFA, paraffin unless stated otherwise. embedded and microtome-sectioned in 5 μm slices. Spinal cord sections were treated to remove paraffin by two steps of 5 min each in xylene and descending alcohols (100%, 95%, 70%, and 50%, 3 min ASSOCIATED CONTENT each) until distilled water, then destined to immunofluorescence analysis. Sections were treated for antigen retrieval with buffer-citrate * sı Supporting Information (citric acid 10 mM with pH = 6, Tween 0.05%; Sigma, USA) at 95 °C The Supporting Information is available free of charge at for 20 min, then cooled, washed in PBS thrice, and permeabilized by https://pubs.acs.org/doi/10.1021/acsnano.2c06609. FBS 5% and Triton 0.3% in PBS, 30 min at RT. Slides were then washed thrice with PBS and labeled at 4 °C overnight with primary Cytokines and chemokines measurements; LY uptake; s- monoclonal antibody rat antimouse GFAP (clone 2.2B10), to detect GO effect on peripheral immune system; genotoxicity astrocytes, with polyclonal antibody rabbit antimouse Iba1, to detect evaluation of s-GO impact; supplementary methods for microglia or with polyclonal antibody NeuN to label neuronal nuclei. UV−vis spectroscopy, X-ray photoemission spectrosco- Afterward, sections were PBS washed thrice and labeled 2 h, RT, in py, TEM analysis on mice spinal cord, Raman the dark, with appropriate secondary antibodies, including Alexa Fluor spectroscopy analysis; lymph node cells and splenocytes 488 goat antirat IgG (H+L) or Alexa Fluor 594 donkey antirabbit IgG proliferation (PDF) (H+L), both from Thermo Fisher Scientific, USA. All antibodies (Thermo Fisher Scientific, USA) were used 1:500 diluted in permeabilizing solution. After secondary antibody, spinal cord sections were PBS washed three times and mounted using mounting AUTHOR INFORMATION medium containing DAPI to detect cell nuclei (Prolong Gold with Corresponding Authors DAPI, Life Technologies, USA). Clara Ballerini − Dipartimento di Medicina Sperimentale e Analysis of Neurons, Astrogliosis, and Microglial Activa- tion. Astrogliosis in spinal cord gray matter and neuronal state were Clinica, University of Florence, 50139 Florence, Italy; evaluated by manual count of GFAP+ (astrocytes) or NeuN+ Email: clara.ballerini@unifi.it (neurons) using ImageJ software (NIH, USA) cells in slide fields at Giada Cellot − International School for Advanced Studies 40× magnification. For each mice group, a range of 5−10 random (SISSA/ISAS), 34136 Trieste, Italy; orcid.org/0000- fields per slide was evaluated (each slide field is equal to an area of 0001-9198-8402; Email: cellot@sissa.it 1975 https://doi.org/10.1021/acsnano.2c06609 ACS Nano 2023, 17, 1965−1978 ACS Nano www.acsnano.org Article Laura Ballerini − International School for Advanced Studies lucifer yellow; CBX, carbenoxolone; GJ, gap-junction; CI, (SISSA/ISAS), 34136 Trieste, Italy; orcid.org/0000- cross-correlation index; QN, quinine 0001-8420-0787; Email: laura.ballerini@sissa.it Authors REFERENCES Giuseppe Di Mauro − International School for Advanced (1) Gilhus, N. E.; Deuschl, G. Neuroinflammation � a Common Studies (SISSA/ISAS), 34136 Trieste, Italy Thread in Neurological Disorders. Nat. Rev. Neurol. 2019, 15 (8), Roberta Amoriello − International School for Advanced 429−430. Studies (SISSA/ISAS), 34136 Trieste, Italy; Dipartimento di (2) Tansey, M. G.; Goldberg, M. S. Neuroinflammation in Parkinson’s Disease: Its Role in Neuronal Death and Implications Medicina Sperimentale e Clinica, University of Florence, for Therapeutic Intervention. Neurobiol. Dis. 2010, 37 (3), 510−518. 50139 Florence, Italy (3) Bjelobaba, I.; Savic, D.; Lavrnja, I. Multiple Sclerosis and Neus Lozano − Catalan Institute of Nanoscience and Neuroinflammation: The Overview of Current and Prospective Nanotechnology (ICN2), 08193 Barcelona, Spain; Therapies. Curr. Pharm. Des. 2017, 23 (5), 693−730. orcid.org/0000-0002-9026-1743 (4) Leng, F.; Edison, P. Neuroinflammation and Microglial Alberto Carnasciali − Dipartimento di Medicina Sperimentale Activation in Alzheimer Disease: Where Do We Go from Here? e Clinica, University of Florence, 50139 Florence, Italy Nat. Rev. Neurol. 2021, 17 (3), 157−172. Daniele Guasti − Dipartimento di Medicina Sperimentale e (5) Jayaraj, R. L.; Azimullah, S.; Beiram, R.; Jalal, F. Y.; Rosenberg, Clinica, University of Florence, 50139 Florence, Italy G. A. Neuroinflammation: Friend and Foe for Ischemic Stroke. J. Maurizio Becucci − Dipartimento di Chimica “Ugo Schiff”, Neuroinflammation 2019, 16 (1), 142. (6) Dinet, V.; Petry, K. G.; Badaut, J.Brain−Immune Interactions DICUS, University of Florence, 50139 Florence, Italy and Neuroinflammation After Traumatic Brain Injury. Front. Neuro- Kostas Kostarelos − Catalan Institute of Nanoscience and sci.2019, 13. DOI: 10.3389/fnins.2019.01178. Nanotechnology (ICN2), 08193 Barcelona, Spain; (7) Magistretti, P. J.; Allaman, I. Lactate in the Brain: From Nanomedicine Lab, and Faculty of Biology, Medicine & Metabolic End-Product to Signalling Molecule. Nat. Rev. Neurosci. Health, The National Graphene Institute, University of 2018, 19 (4), 235−249. Manchester, Manchester M13 9PL, United Kingdom; (8) Makarenkova, H. P.; Shestopalov, V. I. The Role of Pannexin orcid.org/0000-0002-2224-6672 Hemichannels in Inflammation and Regeneration. Front. Physiol. 2014, 5, 63. Complete contact information is available at: (9) Bernaus, A.; Blanco, S.; Sevilla, A. Glia Crosstalk in https://pubs.acs.org/10.1021/acsnano.2c06609 Neuroinflammatory Diseases. Front. Cell. Neurosci. 2020, 14, 209. (10) Craft, J. M.; Watterson, D. M.; Van Eldik, L. J. Neuro- Author Contributions inflammation: A Potential Therapeutic Target. Expert Opin. Ther. # GDM and RA contributed equally. GDM performed spinal Targets 2005, 9 (5), 887−900. slice experiments (imaging, confocal microscopy, immuno- (11) Yang, Y.; Asiri, A. M.; Tang, Z.; Du, D.; Lin, Y. Graphene Based fluorescence) and related analysis; RA performed EAE Materials for Biomedical Applications. Mater. Today 2013, 16 (10), experiments and related clinical and histological analysis; NL 365−373. prepared and characterized the materials; AC performed blind (12) Kostarelos, K.; Novoselov, K. S. Exploring the Interface of evaluation of EAE animals and contributed to the first in vivo Graphene and Biology. Science 2014, 344 (6181), 261−263. tests; DG performed TEM microscopy and analysis; MB (13) Kitko, K. E.; Zhang, Q. Graphene-Based Nanomaterials: From Production to Integration With Modern Tools in Neuroscience. performed Raman spectroscopy; GC performed the electro- Front. Syst. Neurosci. 2019, 13, 26. physiological experiments; KK design the s-GO experimental (14) Tupone, M. G.; Panella, G.; d’Angelo, M.; Castelli, V.; Caioni, strategy; CB design the EAE experimental strategy; KK, CB, G.; Catanesi, M.; Benedetti, E.; Cimini, A. An Update on Graphene- and LB conceived the study; GC, CB, and LB designed the Based Nanomaterials for Neural Growth and Central Nervous System experiments, interpreted the results, and wrote the manuscript. Regeneration. Int. J. Mol. Sci. 2021, 22 (23), 13047. Funding (15) Singh, D. P.; Herrera, C. E.; Singh, B.; Singh, S.; Singh, R. K.; This work received funding from the European Union Horizon Kumar, R. Graphene Oxide: An Efficient Material and Recent Approach for Biotechnological and Biomedical Applications. Mater. 2020 Research and Innovation Programme under Grant Sci. Eng. C Mater. Biol. Appl. 2018, 86, 173−197. Agreement No. GrapheneCore3 (881603). The ICN2 is (16) Esmaeili, Y.; Bidram, E.; Zarrabi, A.; Amini, A.; Cheng, C. funded by the CERCA program, Generalitat de Catalunya, Graphene Oxide and Its Derivatives as Promising In-Vitro Bio- and is supported by the Severo Ochoa Center of Excellence Imaging Platforms. Sci. Rep. 2021, 11 (1), 9071. program by the Spanish Research Agency (AEI, Grant No. (17) Cellot, G.; Franceschi Biagioni, A.; Ballerini, L. Nanomedicine SEV-2017−0706). and Graphene-Based Materials: Advanced Technologies for Potential Notes Treatments of Diseases in the Developing Nervous System. Pediatr. The authors declare no competing financial interest. Res. 2022, 92 (1), 71−79. (18) Rauti, R.; Medelin, M.; Newman, L.; Vranic, S.; Reina, G.; Bianco, A.; Prato, M.; Kostarelos, K.; Ballerini, L. Graphene Oxide ACKNOWLEDGMENTS Flakes Tune Excitatory Neurotransmission in Vivo by Targeting We would like to acknowledge Dr. M. Arellano and A. Hippocampal Synapses. Nano Lett. 2019, 19 (5), 2858−2870. Karakasidi for partial contributions in the analysis of the GO (19) Cellot, G.; Vranic, S.; Shin, Y.; Worsley, R.; Rodrigues, A. F.; material and characterization data. We are grateful to Prof. A. Bussy, C.; Casiraghi, C.; Kostarelos, K.; McDearmid, J. R. Graphene Nistri for critical revision of the manuscript. Oxide Nanosheets Modulate Spinal Glutamatergic Transmission and Modify Locomotor Behaviour in an in Vivo Zebrafish Model. Nanoscale Horiz. 2020, 5 (8), 1250−1263. ABBREVIATIONS (20) Musto, M.; Parisse, P.; Pachetti, M.; Memo, C.; Di Mauro, G.; CKs, cytokines;; GO, graphene oxide; CNS, central nervous Ballesteros, B.; Lozano, N.; Kostarelos, K.; Casalis, L.; Ballerini, L. system; HC, hemichannel; s-GO, small graphene oxide; LY, Shedding Plasma Membrane Vesicles Induced by Graphene Oxide 1976 https://doi.org/10.1021/acsnano.2c06609 ACS Nano 2023, 17, 1965−1978 ACS Nano www.acsnano.org Article Nanoflakes in Brain Cultured Astrocytes. Carbon 2021, 176, 458− (37) Loret, T.; de Luna, L. A. V.; Fordham, A.; Arshad, A.; Barr, K.; 469. Lozano, N.; Kostarelos, K.; Bussy, C. Innate but Not Adaptive (21) Franceschi Biagioni, A.; Cellot, G.; Pati, E.; Lozano, N.; Immunity Regulates Lung Recovery from Chronic Exposure to Ballesteros, B.; Casani, R.; Coimbra, N. C.; Kostarelos, K.; Ballerini, L. Graphene Oxide Nanosheets. Adv. Sci. 2022, 9 (11), No. e2104559. Graphene Oxide Prevents Lateral Amygdala Dysfunctional Synaptic (38) Schwartz, M.; Butovsky, O.; Brück, W.; Hanisch, U.-K. Plasticity and Reverts Long Lasting Anxiety Behavior in Rats. Microglial Phenotype: Is the Commitment Reversible? Trends Biomaterials 2021, 271, 120749. Neurosci. 2006, 29 (2), 68−74. (22) Fabbri, R.; Saracino, E.; Treossi, E.; Zamboni, R.; Palermo, V.; (39) Chhor, V.; Le Charpentier, T.; Lebon, S.; Oré, M.-V.; Celador, Benfenati, V. Graphene Glial-Interfaces: Challenges and Perspectives. I. L.; Josserand, J.; Degos, V.; Jacotot, E.; Hagberg, H.; Sävman, K.; Nanoscale 2021, 13 (8), 4390−4407. Mallard, C.; Gressens, P.; Fleiss, B. Characterization of Phenotype (23) Avossa, D.; Rosato-Siri, M. D.; Mazzarol, F.; Ballerini, L. Spinal Markers and Neuronotoxic Potential of Polarised Primary Microglia Circuits Formation: A Study of Developmentally Regulated Markers in Vitro. Brain Behav. Immun. 2013, 32, 70−85. in Organotypic Cultures of Embryonic Mouse Spinal Cord. (40) Evans, M. C.; Couch, Y.; Sibson, N.; Turner, M. R. Neuroscience 2003, 122 (2), 391−405. Inflammation and Neurovascular Changes in Amyotrophic Lateral (24) Furlan, F.; Taccola, G.; Grandolfo, M.; Guasti, L.; Arcangeli, A.; Sclerosis. Mol. Cell. Neurosci. 2013, 53, 34−41. Nistri, A.; Ballerini, L. ERG Conductance Expression Modulates the (41) Verkhratsky, A.; Kettenmann, H. Calcium Signalling in Glial Excitability of Ventral Horn GABAergic Interneurons That Control Cells. Trends Neurosci. 1996, 19 (8), 346−352. Rhythmic Oscillations in the Developing Mouse Spinal Cord. J. (42) Guerra-Gomes, S.; Sousa, N.; Pinto, L.; Oliveira, J. F. Neurosci. 2007, 27 (4), 919−928. Functional Roles of Astrocyte Calcium Elevations: From Synapses (25) Panattoni, G.; Amoriello, R.; Memo, C.; Thalhammer, A.; to Behavior. Front. Cell. Neurosci. 2018, 11, 427. Ballerini, C.; Ballerini, L. Diverse Inflammatory Threats Modulate (43) Salter, M. W.; Hicks, J. L. ATP-Evoked Increases in Astrocytes Ca2+ Signaling via Connexin43 Hemichannels in Organo- Intracellular Calcium in Neurons and Glia from the Dorsal Spinal typic Spinal Slices. Mol. Brain 2021, 14 (1), 159. Cord. J. Neurosci. 1994, 14 (3), 1563−1575. (26) Giacco, V.; Panattoni, G.; Medelin, M.; Bonechi, E.; Aldinucci, (44) Scemes, E.; Suadicani, S. O.; Spray, D. C. Intercellular A.; Ballerini, C.; Ballerini, L. Cytokine Inflammatory Threat, but Not Communication in Spinal Cord Astrocytes: Fine Tuning between LPS One, Shortens GABAergic Synaptic Currents in the Mouse Gap Junctions and P2 Nucleotide Receptors in Calcium Wave Spinal Cord Organotypic Cultures. J. Neuroinflammation 2019, 16 Propagation. J. Neurosci. 2000, 20 (4), 1435−1445. (1), 127. (45) Fabbro, A.; Pastore, B.; Nistri, A.; Ballerini, L. Activity- (27) Medelin, M.; Giacco, V.; Aldinucci, A.; Castronovo, G.; Independent Intracellular Ca2+ Oscillations Are Spontaneously Bonechi, E.; Sibilla, A.; Tanturli, M.; Torcia, M.; Ballerini, L.; Generated by Ventral Spinal Neurons during Development in Vitro. Cozzolino, F.; Ballerini, C. Bridging Pro-Inflammatory Signals, Cell Calcium 2007, 41 (4), 317−329. Synaptic Transmission and Protection in Spinal Explants in Vitro. (46) Medelin, M.; Rancic, V.; Cellot, G.; Laishram, J.; Mol. Brain 2018, 11 (1), 3. Veeraraghavan, P.; Rossi, C.; Muzio, L.; Sivilotti, L.; Ballerini, L. (28) Fogal, B.; Hewett, S. J. Interleukin-1β: A Bridge between Altered Development in GABA Co-Release Shapes Glycinergic Inflammation and Excitotoxicity? J. Neurochem. 2008, 106 (1), 1−23. Synaptic Currents in Cultured Spinal Slices of the SOD1G93A (29) Probert, L. TNF and Its Receptors in the CNS: The Essential, Mouse Model of Amyotrophic Lateral Sclerosis. J. Physiol. 2016, 594 the Desirable and the Deleterious Effects. Neuroscience 2015, 302, 2− 22. (13), 3827−3840. (30) Baxter, A. G. The Origin and Application of Experimental (47) Stout, C. E.; Costantin, J. L.; Naus, C. C. G.; Charles, A. C. Autoimmune Encephalomyelitis. Nat. Rev. Immunol. 2007, 7 (11), Intercellular Calcium Signaling in Astrocytes via ATP Release through 904−912. Connexin Hemichannels *. J. Biol. Chem. 2002, 277 (12), 10482− (31) Ballerini, C.Experimental Autoimmune Encephalomyelitis. In 10488. T-Helper Cells: Methods and Protocols; Annunziato, F., Maggi, L., (48) Même, W.; Calvo, C.-F.; Froger, N.; Ezan, P.; Amigou, E.; Mazzoni, A., Eds.; Springer US: New York, 2021; pp 375−384. Koulakoff, A.; Giaume, C. Proinflammatory Cytokines Released from DOI: 10.1007/978-1-0716-1311-5_27. Microglia Inhibit Gap Junctions in Astrocytes: Potentiation by β- (32) Rodrigues, A. F.; Newman, L.; Lozano, N.; Mukherjee, S. P.; Amyloid. FASEB J. 2006, 20 (3), 494−496. Fadeel, B.; Bussy, C.; Kostarelos, K. A Blueprint for the Synthesis and (49) de Groot, J. R.; Veenstra, T.; Verkerk, A. O.; Wilders, R.; Smits, Characterisation of Thin Graphene Oxide with Controlled Lateral J. P. P.; Wilms-Schopman, F. J. G.; Wiegerinck, R. F.; Bourier, J.; Dimensions for Biomedicine. 2D Mater. 2018, 5 (3), No. 035020. Belterman, C. N. W.; Coronel, R.; Verheijck, E. E. Conduction (33) Rauti, R.; Lozano, N.; León, V.; Scaini, D.; Musto, M.; Rago, I.; Slowing by the Gap Junctional Uncoupler Carbenoxolone. Cardiovasc. Ulloa Severino, F. P.; Fabbro, A.; Casalis, L.; Vázquez, E.; Kostarelos, Res. 2003, 60 (2), 288−297. K.; Prato, M.; Ballerini, L. Graphene Oxide Nanosheets Reshape (50) Sun, J.-D.; Liu, Y.; Yuan, Y.-H.; Li, J.; Chen, N.-H. Gap Synaptic Function in Cultured Brain Networks. ACS Nano 2016, 10 Junction Dysfunction in the Prefrontal Cortex Induces Depressive-like (4), 4459−4471. Behaviors in Rats. Neuropsychopharmacology 2012, 37 (5), 1305− (34) Secomandi, N.; Franceschi Biagioni, A.; Kostarelos, K.; Cellot, 1320. G.; Ballerini, L. Thin Graphene Oxide Nanoflakes Modulate (51) Verselis, V. K.; Srinivas, M. Connexin Channel Modulators and Glutamatergic Synapses in the Amygdala Cultured Circuits: Their Mechanisms of Action. Neuropharmacology 2013, 75, 517−524. Exploiting Synaptic Approaches to Anxiety Disorders. Nanomedicine (52) Denic, A.; Wootla, B.; Pirko, I.; Mangalam, A.Chapter 11 - 2020, 26, 102174. Pathophysiology of Experimental Autoimmune Encephalomyelitis. In (35) Vranic, S.; Rodrigues, A. F.; Buggio, M.; Newman, L.; White, Multiple Sclerosis; Minagar, A., Ed.; Academic Press: San Diego, 2016; M. R. H.; Spiller, D. G.; Bussy, C.; Kostarelos, K. Live Imaging of pp 249−280. DOI: 10.1016/B978-0-12-800763-1.00011-7. Label-Free Graphene Oxide Reveals Critical Factors Causing (53) Lassmann, H. Axonal and Neuronal Pathology in Multiple Oxidative-Stress-Mediated Cellular Responses. ACS Nano 2018, 12 Sclerosis: What Have We Learnt from Animal Models. Exp. Neurol. (2), 1373−1389. 2010, 225 (1), 2−8. (36) Rodrigues, A. F.; Newman, L.; Jasim, D.; Mukherjee, S. P.; (54) Ballerini, C.; Baldi, G.; Aldinucci, A.; Maggi, P. Nanomaterial Wang, J.; Vacchi, I. A.; Ménard-Moyon, C.; Bianco, A.; Fadeel, B.; Applications in Multiple Sclerosis Inflamed Brain. J. Neuroimmune Kostarelos, K.; Bussy, C. Size-Dependent Pulmonary Impact of Thin Pharmacol. 2015, 10 (1), 1−13. Graphene Oxide Sheets in Mice: Toward Safe-by-Design. Adv. Sci. (55) Bennett, J.; Basivireddy, J.; Kollar, A.; Biron, K. E.; Reickmann, 2020, 7 (12), 1903200. P.; Jefferies, W. A.; McQuaid, S. Blood−Brain Barrier Disruption and 1977 https://doi.org/10.1021/acsnano.2c06609 ACS Nano 2023, 17, 1965−1978 ACS Nano www.acsnano.org Article Enhanced Vascular Permeability in the Multiple Sclerosis Model EAE. Voskuhl, R. R. Cell-Specific and Region-Specific Transcriptomics in J. Neuroimmunol. 2010, 229 (1), 180−191. the Multiple Sclerosis Model: Focus on Astrocytes. Proc. Natl. Acad. (56) Rossi, B.; Constantin, G. Live Imaging of Immune Responses in Sci. U. S. A. 2018, 115 (2), E302−E309. Experimental Models of Multiple Sclerosis. Front. Immunol. 2016, 7, (73) Borggrewe, M.; Grit, C.; Vainchtein, I. D.; Brouwer, N.; 506. Wesseling, E. M.; Laman, J. D.; Eggen, B. J. L.; Kooistra, S. M.; (57) Mandolesi, G.; Gentile, A.; Musella, A.; Fresegna, D.; De Vito, Boddeke, E. W. G. M. Regionally Diverse Astrocyte Subtypes and F.; Bullitta, S.; Sepman, H.; Marfia, G. A.; Centonze, D. Synaptopathy Their Heterogeneous Response to EAE. Glia 2021, 69 (5), 1140− Connects Inflammation and Neurodegeneration in Multiple Sclerosis. 1154. Nat. Rev. Neurol. 2015, 11 (12), 711−724. (74) Jasim, D. A.; Ménard-Moyon, C.; Bégin, D.; Bianco, A.; (58) Ifergan, I.; Davidson, T. S.; Kebir, H.; Xu, D.; Palacios- Kostarelos, K. Tissue Distribution and Urinary Excretion of Macapagal, D.; Cann, J.; Rodgers, J. M.; Hunter, Z. N.; Pittet, C. L.; Intravenously Administered Chemically Functionalized Graphene Beddow, S.; Jones, C. A.; Prat, A.; Sleeman, M. A.; Miller, S. D. Oxide Sheets. Chem. Sci. 2015, 6 (7), 3952−3964. Targeting the GM-CSF Receptor for the Treatment of CNS (75) Newman, L.; Jasim, D. A.; Prestat, E.; Lozano, N.; de Lazaro, I.; Autoimmunity. J. Autoimmun. 2017, 84, 1−11. Nam, Y.; Assas, B. M.; Pennock, J.; Haigh, S. J.; Bussy, C.; Kostarelos, (59) Musto, M.; Rauti, R.; Rodrigues, A. F.; Bonechi, E.; Ballerini, K. Splenic Capture and In Vivo Intracellular Biodegradation of C.; Kostarelos, K.; Ballerini, L. 3D Organotypic Spinal Cultures: Biological-Grade Graphene Oxide Sheets. ACS Nano 2020, 14 (8), Exploring Neuron and Neuroglia Responses Upon Prolonged 10168−10186. Exposure to Graphene Oxide. Front. Syst. Neurosci. 2019, 13, 1. (76) Peng, G.; Montenegro, M. F.; Ntola, C. N. M.; Vranic, S.; (60) Sun, C.; Wakefield, D. L.; Han, Y.; Muller, D. A.; Holowka, D. Kostarelos, K.; Vogt, C.; Toprak, M. S.; Duan, T.; Leifer, K.; A.; Baird, B. A.; Dichtel, W. R. Graphene Oxide Nanosheets Stimulate Bräutigam, L.; Lundberg, J. O.; Fadeel, B. Nitric Oxide-Dependent Ruffling and Shedding of Mammalian Cell Plasma Membranes. Chem. Biodegradation of Graphene Oxide Reduces Inflammation in the 2016, 1 (2), 273−286. Gastrointestinal Tract. Nanoscale 2020, 12 (32), 16730−16737. (61) Orellana, J. A.; Hernández, D. E.; Ezan, P.; Velarde, V.; (77) Mukherjee, S. P.; Gliga, A. R.; Lazzaretto, B.; Brandner, B.; Bennett, M. V. L.; Giaume, C.; Sáez, J. C. Hypoxia in High Glucose Fielden, M.; Vogt, C.; Newman, L.; Rodrigues, A. F.; Shao, W.; Followed by Reoxygenation in Normal Glucose Reduces the Viability Fournier, P. M.; Toprak, M. S.; Star, A.; Kostarelos, K.; Bhattacharya, of Cortical Astrocytes through Increased Permeability of Connexin 43 K.; Fadeel, B. Graphene Oxide Is Degraded by Neutrophils and the Hemichannels. Glia 2009, 58 (3), 329−343. Degradation Products Are Non-Genotoxic. Nanoscale 2018, 10 (3), (62) Orellana, J. A.; Froger, N.; Ezan, P.; Jiang, J. X.; Bennett, M. V. 1180−1188. L.; Naus, C. C.; Giaume, C.; Sáez, J. C. ATP and Glutamate Released (78) Ali-Boucetta, H.; Bitounis, D.; Raveendran-Nair, R.; Servant, via Astroglial Connexin 43 Hemichannels Mediate Neuronal Death A.; Van den Bossche, J.; Kostarelos, K. Purified Graphene Oxide through Activation of Pannexin 1 Hemichannels. J. Neurochem. 2011, Dispersions Lack in Vitro Cytotoxicity and in Vivo Pathogenicity. 118 (5), 826−840. Adv. Healthc. Mater. 2013, 2 (3), 433−441. (63) Lucchinetti, C.; Brü ck, W.; Parisi, J.; Scheithauer, B.; (79) Mukherjee, S. P.; Lozano, N.; Kucki, M.; Del Rio-Castillo, A. Rodriguez, M.; Lassmann, H. Heterogeneity of Multiple Sclerosis E.; Newman, L.; Vázquez, E.; Kostarelos, K.; Wick, P.; Fadeel, B. Lesions: Implications for the Pathogenesis of Demyelination. Ann. Detection of Endotoxin Contamination of Graphene Based Materials Neurol. 2000, 47 (6), 707−717. Using the TNF-α Expression Test and Guidelines for Endotoxin-Free (64) Shigetomi, E.; Saito, K.; Sano, F.; Koizumi, S. Aberrant Calcium Graphene Oxide Production. PLoS One 2016, 11 (11), No. e0166816. Signals in Reactive Astrocytes: A Key Process in Neurological (80) Ballerini, L.; Galante, M.; Grandolfo, M.; Nistri, A. Generation Disorders. Int. J. Mol. Sci. 2019, 20 (4), 996. of Rhythmic Patterns of Activity by Ventral Interneurones in Rat (65) Soulika, A. M.; Lee, E.; McCauley, E.; Miers, L.; Bannerman, Organotypic Spinal Slice Culture. J. Physiol. 1999, 517 (2), 459−475. P.; Pleasure, D. Initiation and Progression of Axonopathy in (81) Galante, M.; Nistri, A.; Ballerini, L. Opposite Changes in Experimental Autoimmune Encephalomyelitis. J. Neurosci. 2009, 29 Synaptic Activity of Organotypic Rat Spinal Cord Cultures after (47), 14965−14979. Chronic Block of AMPA/kainate or Glycine and GABAA Receptors. (66) Zhu, F.-D.; Hu, Y.-J.; Yu, L.; Zhou, X.-G.; Wu, J.-M.; Tang, Y.; J. Physiol. 2000, 523 (3), 639−651. Qin, D.-L.; Fan, Q.-Z.; Wu, A.-G. Nanoparticles: A Hope for the (82) Cossarizza, A.; Chang, H.-D.; Radbruch, A.; Acs, A.; Adam, D.; Treatment of Inflammation in CNS. Front. Pharmacol. 2021, 12, Adam-Klages, S.; Agace, W. W.; Aghaeepour, N.; Akdis, M.; Allez, M.; 683935. Almeida, L. N.; Alvisi, G.; Anderson, G.; Andrä, I.; Annunziato, F.; (67) Kenison, J. E.; Jhaveri, A.; Li, Z.; Khadse, N.; Tjon, E.; Tezza, Anselmo, A.; Bacher, P.; Baldari, C. T.; Bari, S.; Barnaba, V.; Barros- S.; Nowakowska, D.; Plasencia, A.; Stanton, V. P., Jr; Sherr, D. H.; Martins, J.; Battistini, L.; Bauer, W.; Baumgart, S.; Baumgarth, N.; Quintana, F. J. Tolerogenic Nanoparticles Suppress Central Nervous Baumjohann, D.; et al. Guidelines for the Use of Flow Cytometry and System Inflammation. Proc. Natl. Acad. Sci. U. S. A. 2020, 117 (50), Cell Sorting in Immunological Studies (second Edition). Eur. J. Immunol. 2019, 49 (10), 1457−1973. 32017−32028. (68) Jan, Z.; Mollazadeh, S.; Abnous, K.; Taghdisi, S. M.; Danesh, A.; Ramezani, M.; Alibolandi, M. Targeted Delivery Platforms for the Treatment of Multiple Sclerosis. Mol. Pharmaceutics 2022, 19 (7), 1952−1976. (69) Mohammadpour, R.; Dobrovolskaia, M. A.; Cheney, D. L.; Greish, K. F.; Ghandehari, H. Subchronic and Chronic Toxicity Evaluation of Inorganic Nanoparticles for Delivery Applications. Adv. Drug Delivery Rev. 2019, 144, 112−132. (70) Ivashkevich, A.; Redon, C. E.; Nakamura, A. J.; Martin, R. F.; Martin, O. A. Use of the γ-H2AX Assay to Monitor DNA Damage and Repair in Translational Cancer Research. Cancer Lett. 2012, 327 (1−2), 123−133. (71) Teleanu, D. M.; Chircov, C.; Grumezescu, A. M.; Teleanu, R. I. Neuronanomedicine: An Up-to-Date Overview. Pharmaceutics 2019, 11 (3), 101. (72) Itoh, N.; Itoh, Y.; Tassoni, A.; Ren, E.; Kaito, M.; Ohno, A.; Ao, Y.; Farkhondeh, V.; Johnsonbaugh, H.; Burda, J.; Sofroniew, M. V.; 1978 https://doi.org/10.1021/acsnano.2c06609 ACS Nano 2023, 17, 1965−1978