n-3 Docosapentaenoic acid-derived protectin D1 promotes resolution of neuroinflammation and arrests epileptogenesis

Frigerio et al. report that pro-resolving mechanisms of neuroinflammation are dysregulated during epileptogenesis, thereby promoting a persistent neuroinflammatory response that contributes to seizure generation and cognitive deficits. Boosting endogenous resolution responses by administering a specific lipid mediator improves disease outcomes in a murine epilepsy model, suggesting a novel treatment avenue.

CT values were obtained using manual threshold and baseline, analyzed using the 2 -ΔΔCt method and normalized using geometric mean of 3 independent house-keeping genes (Mfsd5,Brap,Bcl2l13); Results are shown in Figures 1, 4 and Suppl. Fig. 1 and Suppl. Fig. 2.

Pilocarpine injection in mice
Adult male NMRI mice were used. Pilocarpine mice were generated as described previously (Mazzuferi et al., 2012). Pilocarpine (300 mg/kg, Sigma-Aldrich) was intraperitoneally (i.p.) injected 30 min after i.p. administration of 1 mg/kg of N-methylscopolamine bromide (Sigma-Aldrich). Behavioral endpoints, such as the time to onset and the duration of SE as well as the survival rate were assessed. SE typically appeared within the first hour after pilocarpine injection (59.3 ± 2.6 min) and was characterized by continuous stage 3 to 5 motor seizures (Racine, 1972) without regaining consciousness (unresponsiveness to any environmental stimuli) together with loss of postural control. This definition is consistent with that being commonly used in the rat pilocarpine model (Turski et al., 1983;Cavalheiro et al., 1987). SE motor episodes were stopped after 2 h with an acute i.p. injection of diazepam (10 mg/kg). Sham mice were subjected to a similar schedule of injections, but with saline instead of pilocarpine. All mice injected with pilocarpine and their controls were killed 72 h post-SE for hippocampi dissection and RTqPCR analysis.

Novel object recognition test
We employed this behavioral test (NORT) to assess the ability of rodents to recognize a set of novel objects in an otherwise familiar environment since this behavior is taken as a measure of recognition memory (Balducci et al., 2010). This task is predominantly associated with limbic cortex function although the 24 h inter-trial interval chosen from familiarization to test phase also involves hippocampal activity (Balducci et al., 2010;Mazarati et al., 2011). Recognition memory was tested in mice exposed to SE and treated i.c.v. for 4 days with 200 ng/μl PD1 n-3DPA ME (n=11) or saline (n=12). Control mice (sham) were similarly implanted, injected with vehicle but not exposed to SE (n=18). Each mouse was weighted between 8:00 and 9:00 a. The test began on the next day with the familiarization phase, when mice were placed into the open field for 10 min in the presence of two identical objects positioned in internal nonadjacent squares.
The following objects were randomly used: black plastic cylinders (4×5 cm); transparent scintillation vials with white cups (3×6 cm); metal cubes (3×5 cm). Cumulative exploration time of both objects and of each object separately was recorded. Exploration was defined as sniffing, touching, and stretching the head toward the object at a distance of not more than 2 cm. Twenty four hours after familiarization, the recognition phase of the test was performed: mice were placed for 10 min in the open field which contained one object presented during the familiarization phase (familiar object, F) and a novel unfamiliar object (N). The time spent exploring N vs F, as well as cumulative exploration time (i.e. novel+familiar, N+F) was recorded. As the recognition phase was performed 24 h after the familiarization phase, the procedure can be regarded as a test of long-term memory. Novel object recognition was quantified using the discrimination index (N-F/N+F): the difference between the time spent exploring the novel and the familiar objects (N-F) divided by the sum of total exploration time (N+F) (Okun et al., 2010). At the end of NORT, all mice were sacrificed for histopathological analysis (see below).

Immunohistochemistry and double-immunostaining
Mice were sacrificed 72 h after SE onset for histological analysis (n=8) (Fig. 1). Control mice (sham) were implanted with electrodes and guide cannulae and injected with vehicles but were not exposed to SE (n=8). Mice were deeply anaesthetized by injecting ketamine (75 mg/kg, i.p.) and medetomidine (0.5 mg/kg, i.p.), then perfused via ascending aorta with 50 mM ice-cold PBS, pH 7.4 followed by chilled 4% paraformaldehyde (Merck, Darmstadt, Germany, #104005) in PBS. The brains were post-fixed for 90 min at 4° C, then transferred to 20% sucrose in PBS for 24 h at 4° C.
Then, the brains were immersed in -45°C isopentane for 3 min and stored at -80°C until assayed.
Serial coronal sections were cut on a cryostat throughout the septo-temporal extension of the hippocampus (Franklin and Paxinos, 2008) (-0.94 to -3.64 mm from bregma). Four series of 16 sections each brain were prepared and in 3 series the slices were stained as follows: the 1 st slice for IL-1β, the 2 nd for TNF-α, the 4 th for ALXR and the 5 th for ChemR23. Three slices per mice were used for each marker. The same anatomical structures were retained within each series of sections to be compared.

Human subjects
The cases included in this study were obtained from the archives of the Departments of Neuropathology of the Academic Medical Center (AMC, Amsterdam, The Netherlands) and the VU University medical center (VUmc, Amsterdam, The Netherlands). A total of 7 hippocampal specimens (removed from patients with temporal lobe epilepsy with hippocampal sclerosis (TLE-HS) undergoing surgery for drug-resistant epilepsy) and 7 hippocampal specimens obtained at autopsy from patients who died after SE were examined. Control material was obtained at autopsy from 6 age-matched control patients, without a history of seizures or other neurological diseases.
All autopsies were performed within 24 h after death. Tissue was obtained and used in accordance with the Declaration of Helsinki and the AMC Research Code provided by the Medical Ethics Committee. All cases were reviewed independently by two neuropathologists and the classification of hippocampal sclerosis was based on analysis of microscopic examination as described by the International League Against Epilepsy (Blumcke et al., 2013). The clinical information of each patient is reported in Suppl. Table 1.

ALXR/FPR2 and ChemR23/ERV1 immunohistochemistry in human tissue
Human brain tissue was fixed in 10% buffered formalin and embedded in paraffin. Tissue was sectioned at 5 µm, mounted on pre-coated glass slides (Star Frost, Waldemar Knittel, Braunschweig, Germany) and processed for immunohistochemical staining (Suppl. Fig. 2E,F).

Histological analysis and quantification of neuronal cell loss, neurogenesis and glia activation in mouse brain
Analyses were performed in the hippocampus ipsilateral to the injected amygdala and in the injected amygdala since in this epilepsy model the histopathology is mostly present in the injected hemisphere (Mouri et al., 2008;Iori et al., 2017). Mice were randomly selected in the various experimental groups for histological analysis (Suppl . Fig 4 and Suppl. 5).
Nissl or Fluoro-Jade (FJ) staining were performed to assess neuronal cell loss and degenerating neurons, respectively. Immunohistochemistry was done to analyse astrocytes (S100) and microglia (Iba-1), and neurogenesis was assessed using doublecortin (DCX). Serial coronal sections were cut on a cryostat throughout the septo-temporal extension of the mouse hippocampus (Franklin and Paxinos, 2008) (-0.94 to -3.64 mm from bregma). Four series of 16 sections each were prepared and in each series the slices were stained as follows: the 1 st slice Nissl, the 2 nd FJ, the 3 rd DCX, the 4 th S100 and the 5 th Iba-1. Immunohistochemical analysis of 3 brain slices per mouse (-1.34, -1.46, -1.58 mm from bregma) and quantification procedures were performed by two independent expert investigators blind to the identity of the samples.
Neuronal cell loss was quantified by reckoning the number of Nissl-stained neurons in the basolateral amygdala and the hilar interneurons using a digitized image of the whole hemisphere captured at 20X magnification (Virtual Slider Microscope; Olympus, Germany). Neuronal cells in CA1, CA3 pyramidal layers were too dense to allow for a sound quantification by cell counting therefore we measured the Nissl-positive area as previously described (Iori et al., 2013). Highpower non-overlapping fields of the whole hippocampus (20X magnification; Olympus) were acquired to measure the total area (µm 2 ) occupied by Nissl-stained neurons along the CA1 and CA3 pyramidal cell layers which reflects neuronal density in each region, using ImageJ software.
Data obtained in each image within the same hippocampal subfield were added together providing one single value per slice in each mouse. Data obtained in each of the 3 slices per brain were averaged, providing a single value for each brain, and this value was used for statistical analysis.
Degenerating neurons were identified by FJ labeling and counted as previously described (Schmued et al., 1997;Ravizza and Vezzani, 2006). Briefly, sections were dried in ethanol (100%, 75% and 50%) and rehydrated in distilled water. Then, they were incubated in 0.06% potassium permanganate, washed in distilled water and transferred to 0.001% FJ staining solution. Sections were then rinsed in distilled water, mounted onto gelatin-coated slides, dried, immersed in xylene and coverslipped. High-power fields (20X magnification; Olympus) along the CA1 and CA3 pyramidal cell layers, the hilus, the basolateral amygdala, the somatosensory/perirhinal cortices were acquired. Nissl-positive cells and FJ-positive neurons were marked and an automated cell count was generated using Fiji software. Data obtained in each slice/area/brain were averaged providing a single value per mouse, and this value was used for statistical analysis. S100, Iba-1 and DCX immunostaining was carried out as previously described (Ravizza et al., 2008;Filibian et al., 2012;Iori et al., 2013). Lack of immunostaining was observed when slices were incubated with the primary antibodies preabsorbed with an excess of the corresponding peptides, or without the primary antibodies. S100 or Iba-1-immunostained area was measured in the whole hippocampus (20X magnification) using ImageJ software. The area was expressed as positive pixels/total assessed pixels; the percentage area with the specific staining was used for subsequent statistical analysis.
The quantification of the total number of S100-immunoreactive astrocytes was carried out using an image of the whole hippocampus captured at 20X magnification (Virtual Slider Microscope; Olympus, Germany) by an investigator who identified the cells; then an automated cell count was generated using ImageJ software.
DCX-immunostained cells were counted using an image of the hilus and the surrounding superior and inferior granule cell layer captured at 20X magnification (Virtual Slider Microscope; Olympus, Germany) as previously described (Pascente et al., 2016).
Data obtained in each mouse hippocampus were averaged, thus providing a single value for each mouse, and this value was used for the statistical analysis. Although this cell counting method has some limitations as compared to designed-based stereological analysis (Schmitz and Hof, 2005) the occurrence of any bias in counting should similarly affect sham and experimental mice since these samples underwent the same procedure in parallel. co-localized also with IL-1β (red). The double-immunostaining of ChemR23/ERV1 with IL-1β was not performed since the two primary antibodies were raised in the same host specie (goat). Colocalization signal is depicted in yellow. Scale bar 50 μm.

Supplementary Figures and Tables
While ALXR/FPR2 staining was diffusely increased in the various hippocampal layers (not shown), ChemR23/ERV1 was predominantly induced in strata radiatum and moleculare of CA1.
In sham mice ALXR and ChemR23/ERV1 immunoreactivity was also detected in NeuN-positive pyramidal neurons (Suppl. Fig. 2B,D), in granule cells (not shown) and scattered interneurons (Suppl. Fig. 2B,D). This staining was apparently reduced after SE likely reflecting neuronal cell loss (not shown).
Human tissue. Panel E shows ALXR/FPR2 and ChemR23/ERV1 immunostaining in CA1 area from autoptic control tissue (n=6) and patients who died between 1 and 49 days post-SE (n=7) or patients affected by chronic epilepsy (temporal lobe epilepsy with hippocampal sclerosis; TLE-HS, n=7).

chronic epileptic mice
Panel A depicts representative Nissl-stained sections showing neurons in CA1 and CA3 pyramidal layers, the hilus and the amygdala of epileptic mice treated with saline or PD1 n-3DPA -ME during epileptogenesis, and in sham mice (not exposed to SE). Mice were killed at the end of EEG recordings ( Figure 6). Panel B shows the quantification of neuronal cell loss in mice randomly selected in the various experimental groups. Data are presented as box-and-whisker plots depicting median, interquartile interval, minimum and maximum (n=6-7) expressed as % change vs respective SHAM values. Statistical analysis was done using absolute values. *p<0.05, **p<0.01 vs SHAM by Kruskal Wallis test with Dunn's post-hoc correction. Scale bar in panel A: 100 µm.