Profiling the Expression of Endoplasmic Reticulum Stress Associated Heat Shock Proteins in Animal Epilepsy Models

Journal Pre-proofs Research Article Profiling the expression of endoplasmic reticulum stress associated heat shock proteins in animal epilepsy models Marta Nowakowska, Fabio Gualtieri, Eva-Lotta von Rüden, Florian Hansmann, Wolfgang Baumgärtner, Andrea Tipold, Heidrun Potschka PII: DOI: Reference:

S0306-4522(19)30863-2 https://doi.org/10.1016/j.neuroscience.2019.12.015 NSC 19425

To appear in:

Neuroscience

Received Date: Accepted Date:

6 September 2019 9 December 2019

Please cite this article as: M. Nowakowska, F. Gualtieri, E-L. von Rüden, F. Hansmann, W. Baumgärtner, A. Tipold, H. Potschka, Profiling the expression of endoplasmic reticulum stress associated heat shock proteins in animal epilepsy models, Neuroscience (2019), doi: https://doi.org/10.1016/j.neuroscience.2019.12.015

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2019 Published by Elsevier Ltd on behalf of IBRO.

Profiling the expression of endoplasmic reticulum stress associated heat shock proteins in animal epilepsy models Marta Nowakowska1♠, Fabio Gualtieri1♠, Eva-Lotta von Rüden1, Florian Hansmann2, Wolfgang Baumgärtner2, Andrea Tipold3, and Heidrun Potschka1 1Institute

of Pharmacology, Toxicology, and Pharmacy, Ludwig-Maximilians-University

Munich, Germany 2Department 3Department

of Pathology, University of Veterinary Medicine Hannover, Germany of Small Animal Medicine and Surgery, University of Veterinary Medicine

Hannover, Germany ♠

Both authors contributed equally to this manuscript

Correspondence: Prof. Dr. Heidrun Potschka Institute of Pharmacology, Toxicology, and Pharmacy, Ludwig-Maximilians-University Königinstrasse 16, D-80539 Munich, Germany; Phone: +49 (0) 89/2180 - 2663; Fax: +49 (0) 89/2180 - 16556; E-mail: [email protected]

1

Abstract Unfolded protein response is a signaling cascade triggered by misfolded proteins in the endoplasmic reticulum. Heat shock protein H4 (HSPH4) and A5 (HSPA5) are two chaperoning proteins present within the organelle, which target misfolded peptides during prolonged stress conditions. Epileptogenic insults and epileptic seizures are a notable source of stress on cells. To investigate whether they influence expression of these chaperones, we performed immunohistochemical stainings in brains from rats that experienced a status epilepticus (SE) as a trigger of epileptogenesis and from canine epilepsy patients. Quantification of HSPA5 and HSPH4 revealed alterations in hippocampus and parahippocampal cortex. In rats, SE induced up-regulation of HSPA5 in the piriform cortex and down-regulation of HSPA5 and HSPH4 in the hippocampus. Regionally restricted increases in expression of the two proteins has been observed in the chronic phase with spontaneous recurrent seizures. Confocal microscopy revealed a predominant expression of both proteins in neurons, no expression in microglia and circumscribed expression in astroglia. In canine patients, only up-regulation of HSPH4 expression was observed in Cornu Ammonis 1 region in animals diagnosed with structural epilepsy. This characterization of HSPA5 and HSPH4 expression provided extensive information regarding spatial and temporal alterations of the two proteins during SE-induced epileptogenesis and following epilepsy manifestations. Up-regulation of both proteins implies stress exerted on ER during these disease phases. Taken together suggest a differential impact of epileptogenesis on HSPA5 and HSPH4 expression and indicate them as a possible target for pharmacological modulation of unfolded protein response. Keywords: Chaperone; Endoplasmic Reticulum; Seizure; Unfolded Protein Response; Hippocampus

Abbreviations AMP: adenosine monophosphate CA1: hippocampal Cornu Ammonis 1 CA3: hippocampal Cornu Ammonis 3 CE: coefficient of error CV: coefficient of variation CTR: control DAMPs: damage-associated molecular pattern molecules 2

ER: endoplasmic reticulum HIER: heat-induced epitope retrieval HSP: heat shock protein HSPA1A: HSP isoform A1A (also known as HSP70) HSPH4: HSPs isoform H4 (also known as HYOU/Grp170, ORP150) HSPA5: HSPs isoform A5 (also known as Grp78, BIP) OD: optical density PBS (T): Phosphate buffer saline (-added with 0.05% Tween 20) SE: status epilepticus TLE: temporal lobe epilepsy TBS (T): Tris-buffered saline (-added with 0.05% Tween 20) UPR: unfolded protein response

3

Introduction Heat shock proteins (HSPs) comprise a group of highly conserved, ubiquitous molecules, promoting proper folding and assembling of polypeptides (Stetler et al., 2010). They serve a prominent role in the defense against accumulation of misfolded proteins within a cell. Moreover, they execute other cytoprotective effects, like inhibition of apoptosis, protection of cytoskeletal structures and immune modulation in stress conditions (Benarroch, 2011). Their vast functionality and ubiquitous presence in many cell compartments establishes HSPs as key particles necessary for preservation of homeostasis and renders them putative therapeutic targets (Soti et al., 2005). In eukaryotic cells, synthesis of proteins destined for secretion or membrane anchoring occurs mostly in the endoplasmic reticulum (ER). In the ER proteins are folded into their tertiary and quaternary structures in environmental conditions different from the ones in the cytosol (Braakman & Bulleid, 2011). Molecular perturbations reduce the ER protein-folding capacity, leading to accumulation and aggregation of unfolded proteins – a condition known as ER stress (Jaronen et al., 2014). ER stress triggers the unfolded protein response (UPR), which is an adaptive response responsible for restoring protein homeostasis (Kozutsumi et al., 1988). One of the important elements of the UPR is heat shock protein A5 (HSPA5, also known as 78 kDa glucose-regulated protein [GRP78] or binding immunoglobulin protein [BiP[), which belongs to HSP group A (or HSP70). It is expressed constitutively in the ER where it functions as the main endoplasmic chaperone (Ni et al., 2011). Following ER stress, misfolded proteins commence to aggregate inside the organelle. HSPA5 is bound under physiological conditions to three transmembrane signal transducers. In the presence of misfolded proteins, it is released from membrane proteins into the lumen of the ER. Its dissociation enables the signal transducers to perform their previously inhibited activity, which leads to the activation of the UPR signaling pathways (Timberlake et al., 2018). The subsequent unfolded protein response involves both rapid posttranslational regulation of chaperones already active in the ER as well as slower induction of transcription of genes involved in protein quality control. In the end, it promotes degradation of cytotoxic misfolded proteins and cell survival (Wang et al., 2009). HSPA5 can also be expressed in other cell compartments, like mitochondria, cytosol or nucleus (Ni et al., 2011), as well as on the cell membrane and released into the extracellular space, both, in pathological and in physiological conditions (Delpino & Castelli, 2002; MarinBriggiler et al., 2010; Aksoy et al., 2017). Heat shock protein H4 (HSPH4, also known as 150 kDa oxygen-regulated protein [ORP150] and hypoxia up-regulated protein 1 [HYOU1]) belongs to HSP group H (or HSP110), acting 4

as co-chaperones for HSPA subfamily. Given its ADP-ATP exchange function, it mainly acts as a nucleotide exchange factor for HSPA5 in the ER, where it enhances substrate release in an ATP-dependent manner (Dragovic et al., 2006; Raviol et al., 2006). In addition, this peptide is also an independent chaperone directly involved in UPR, but the regulation of its substrate binding function is different from that of other conventional HSP superfamily members (Behnke et al., 2015). HSPH4 has been identified in the ER of all eukaryotic cells (Rachel et al., 1997) where, in concert with HSPA5, it regulates immunoglobulin folding and assembly (Easton et al., 2000). Alternative locations of this protein comprise also the cytosol and the extracellular space where it acts as a modulator of the immune system (Shaner & Morano, 2007; Zuo et al., 2012). Increased ER stress, with the consequent UPR activation followed by HSPA5 and HSPH4 upregulation, has been linked to brain ischemia (Su & Li, 2016), kainic acid-dependent cell injury (Sokka et al., 2007a) and cerebellar Purkinje cell degeneration (Ni & Lee, 2007). The role played by ER stress following status epilepticus (SE) induced cell death is still debated. Some authors report that ER stress may lead to neuronal death (Yamamoto et al., 2006; Sokka et al., 2007a; Chen et al., 2013; Chen et al., 2014), while others sustain a pro-survival mechanism in response to seizures (Kitao et al., 2001a; Torres-Peraza et al., 2013). An increasing amount of data suggests that prolonged ER stress may initiate intracellular signaling contributing to neurodegeneration in the epileptic brain (Sokka et al., 2007b; Ko et al., 2015). In post-status epilepticus models of temporal lobe epilepsy (TLE), status epilepticus triggers epileptogenesis, a process of alterations in neuronal circuitry occurring from the initial insult on and resulting in manifestation of epilepsy with recurrent spontaneous epileptic seizures. Epileptogenesis can be divided in three main stages: 1) the initial precipitating insult followed by the early post-insult phase, 2) the seizure-free latency phase and 3) the chronic epilepsy phase during which patients exhibit spontaneous recurrent seizures (Maguire, 2016). In the present study, we opted for two very different animal models of TLE. The first one is a rat electrical post-SE model, which has been extensively validated in our group (Keck et al., 2017; van Dijk et al., 2018; Gualtieri et al., 2019) and which allowed us to elucidate alterations in the expression profiles of several stress-associated proteins (Walker et al., 2016; Keck et al., 2018), including a recent publication on a member of the heat shock protein family, HSPA1A, belonging to the group A subfamily as HSPA5 (Gualtieri et al., 2019). The second one is a spontaneous canine clinical model with different etiologies of epilepsy that has already been suggested as a translational bridge between highly standardized rodent models and human clinical studies (Potschka et al., 2013). 5

Understanding whether and to which extent HSPH4 and HSPA5 expression vary during epileptogenesis is essential for a better comprehension of the impact of the ER stress and its aftermaths on the pathophysiology of TLE. Investigating the alterations in both, laboratory and clinical conditions, brings an advantage of in-depth learning of the process in a regulated setting as well as in a less predictable clinical environment.

6

Experimental procedures Animals SE induction and tissue collection in rats The brain samples were acquired from the in-house collection from earlier experiments. Animals, purchased from Harlan Laboratories (Udine, Italy), were housed under controlled environmental conditions (20–24 °C, 45–65% humidity, light cycle from 7:00 a.m.–7:00 p.m.) and received food (Ssniff R/M Haltung, ssniff Spezialdiäten GmbH, Germany) and tap water ad libitum. Female rats were selected due to their lower mortality rate as compared to males in this model (Brandt et al., 2003a). Experimental procedures had been carried out in accordance with the German Animal Welfare Act, the European Communities Council Directive of 22 September 2010 (2010/63/EU), in compliance with the ARRIVE (Animal Research: Reporting of In Vivo Experiments) guidelines, and had been approved by the Committees of the Government of Upper Bavaria (reference numbers: AZ 55.2-1-54-2532-94-11, AZ 55.2-1-542532-011-2015). Every attempt had been made to minimize the number of animals used and to avoid any pain or discomfort. Animals with status epilepticus induction were age-matched with control animals (CTR), which underwent electrode surgical implant without stimulation. Combined stimulationrecording electrodes (Teflon® insulated bipolar stainless steel electrodes, Ø 0.45 mm) were positioned by means of a stereotaxic frame (TSE Systems GmbH, Bad Homburg, Germany) under general anesthesia with chloral hydrate (i.p. 360 mg/kg) in the right anterior basolateral amygdala (AP -2.2 mm, ML +4.7 mm, DV -8.5 mm) with coordinates according to the rat brain atlas of Paxinos and Watson (1998). Meloxicam (Metacam®, Boehringer-Ingelheim, Germany) was administered subcutaneously at a dose of 1 mg/kg 30 min pre- and 24 h postsurgery with additional local administration of bupivacaine (Bupivacaine 0.5%, Jenapharm, Jena, Germany) to the surgical site. Pre- and postoperatively, marbofloxacin (1 mg/kg, Marbocyl FD 1%, Vétoquinol, Ravensburg, Germany) was subcutaneously administered twice a day, starting one day before the implantation until the seventh post-surgical day. Eight weeks after surgery SE was induced as described by Ongerth et al. (2014). Briefly, following a baseline electroencephalogram recording, rats were stimulated continuously via the depth electrode for 25 min (intratrain pulse frequency of 50 Hz, 700 µA peak pulse intensity, 100 ms trains of 1 ms alternating positive and negative square-wave-pulses at a frequency of 2 Hz; stimulator Accupulser, A310C and stimulus isolator A365, World precision Instruments, Berlin, Germany). The self-sustained SE was confirmed by electroencephalography recording 7

right after the stimulation. SE was terminated by a diazepam injection (i.p. 20 mg/kg – Diazepam - Ratiopharm, Germany) 4 h after the beginning of the experiment. Behavioral seizures were scored accordingly to the Racine scale (1972) and SE severity ranked as follows: type 1 - partial SE consisting of non-convulsive seizures, type 2 - partial SE with generalized seizures, during which predominantly partial seizures are interrupted by episodes of generalized convulsive seizures, type 3 - generalized SE consisting of convulsive seizure activity (Brandt et al., 2003b). Only brain samples from animals with a type 3 SE were selected for immunohistochemistry as only these animals reliably develop epilepsy. Brains from 36 female Sprague Dawley rats were used for the analysis. Brain tissue was collected at three different time points during epileptogenesis: i) two days post-SE (early post-insult phase), ii) ten days post-SE (latency phase), iii) eight weeks post-SE (chronic epilepsy phase with spontaneous recurrent seizures). SE and CTR animals for each time point were deeply anesthetized by an intraperitoneal injection of 600 mg/kg pentobarbital (Narcoren®, Merial GmbH, Halbergmoos, Germany) and transcardially perfused with 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4). Brains were post-fixed in 4% paraformaldehyde (24 h, pH 7.4) cryoprotected in a solution of 30% sucrose in 0.01 M Phosphate buffer saline (PBS), and embedded in optimal cutting temperature compound (4583, Electron Microscopy Sciences - USA). Coronal sections (40 μm) were cut on a cryostat (HM 560, Microm – Germany). Serial 40 μm coronal brain sections from each time point were then stored at -20°C for several (6-8) months in a cryoprotectant solution (30% ethylene glycol, 30% glycerin, 10% 0.2M phosphate buffer), and afterwards they were processed for immunohistochemistry. Clinical background and tissue collection in dogs Post mortem brain tissue was taken from an in-house collection (Institute for Pathology of the University of Veterinary Medicine in Hannover) of 48 canine brain samples (age range 2 months – 15 years) as previously described (Rüden et al., 2012). Brains were initially postfixed in 10% formalin for ten days, paraffin embedded and cut in 2-µm thick transverse sections before being mounted on positively charged microscope slides (Superfrost plus, MenzelGläser, Braunschweig, Germany). The sections were stored at room temperature, in conditions enabling protection from light. Every section contained the hippocampus in a range from #1360 to #1660 of the canine brain atlas (http://neurosciencelibrary.org). Dogs were distributed to following groups: i) patient control group (CTRpat) consisting of owner kept dogs without central nervous system diseases (n = 18, age range 2-180 months, mean 70.67  12.58), ii) 8

experimental control group (CTRexp) consisting of dogs without CNS diseases (free of neuropathology and with no general pathological findings in the periphery) but previously used as experimental dogs in parasitology research by the Institute of Parasitology of the University of Veterinary Medicine Hanover, Germany; this group was highly standardized and consisted of Beagle dogs kept under uniform conditions (n = 10, age range 12-16 months, mean 14  0.67), iii) epileptic animals consisting of dogs with an epilepsy diagnosis. This latter group was subclassified by the type of epilepsy [according to the etiology definition given by the international veterinary task force (Berendt et al., 2015)], comprising animals presenting structural epilepsy with identified cerebral pathology (n = 12, age range 30-140 months, mean 81.17  12.29) and animals presenting idiopathic epilepsy, with unknown cause and no identification of structural epilepsy (n = 8, age range 2.5 - 157 months, mean 51.94  18.82). The classification has been based on the clinical diagnosis, such as anamnesis, neurological examination and pathological evaluation of canine patients. All animal experiments were conducted in accordance with the German Animal Welfare Law and all experiments were approved by the local authorities (Niedersächsisches Landesamt für Verbraucherschutz und Lebensmittelsicherheit (LAVES), Oldenburg, Germany, permission numbers: 33.9-42502-05-12A241, 33.19-42502-05-16A044). The owners of dogs included in this study signed an informed consent form. Brain tissue from the archive of the department of Pathology of the University of Veterinary Medicine in Hanover, Germany has been used in accordance to the ethical rules and in compliance with the recommendations of the University of Veterinary Medicine in Hanover, Germany. Immunohistochemistry All antibodies used of this study are reported with relative SciCrunch Research Resource Identifiers (https://scicrunch.org/resources) in Table S1. Rat sections: for both HSPH4 and HSPA5 single immunostainings, free-floating 40-μm thick coronal sections were washed in Tris-buffered saline containing 0.05 % Tween-20 (TBST P9416, Sigma-Aldrich, Germany) at room temperature and a heat-induced epitope retrieval (HIER) step was performed in a pre-heated water bath using sodium citrate buffer (80 °C, 30 min, pH 6.0). Following three further TBST washes, endogenous peroxidase was inhibited by 20 min incubation in 3% H2O2 and in order to avoid non-specific antibody binding we performed a 60 min blocking step (5% goat serum, 2% bovine serum albumin – A4737 SigmaAldrich, Germany). Slices were then incubated overnight (16 h, 4 °C) with either anti-HSPA5 9

(1:5000) or anti-HSPH4 (1:750) antibodies. The following day, sections were rinsed thrice in TBST and incubated at room temperature with biotinylated goat anti-rabbit (for HSPH4: 90 min, 1:1000; for HSPA5: 120 min, 1:1500) diluted in antibody carrier (TBS added with 1% bovine serum albumin, Triton 0.3%, 1% goat serum). After three more washing steps in TBST, brain sections were incubated at room temperature in VECTASTAIN ABC-Peroxidase Kit (1:100, 60 min) and stained using SIGMAFAST™ 3,3′-diaminobenzidine tablets in accordance with manufacturer’s instructions. HSPH4 samples were additionally counterstained with Hemalum solution acidic according to Mayer (Roth T865, Carl Roth, Germany) since cell morphology was not as clear as in the HSPA5 staining. All sections were afterwards mounted on microscope glasses and coverslipped with Entellan® (107960, Merck, Darmstadt, Germany). Negative controls were processed in parallel with omission of the primary antibody. Dog sections: For both HSPH4 and HSPA5 immunostaining, paraffin-embedded 2-µm thick brain sections were initially deparaffinized and rehydrated using ethanol and xylene gradients. Since this is a whole mount reaction, all immunohistochemistry incubations were performed in humidity chambers and in-between washes were carried out in cuvettes. Initial washes, HIER and endogenous peroxidase inhibition were performed as previously described for rat brain tissue. Subsequently, sections were blocked using 0.25% casein (C7078 Sigma-Aldrich, Germany) in TBS and incubated overnight (16 h, 4 °C) with either anti-HSPH4 (1:50) or antiHSPA5 (1:50) primary antibodies suspended in antibody carrier (TBS added with 0.25% casein and 0.1% Tween 20). The following day, sections were rinsed thrice in TBST and incubated at room temperature with biotinylated goat anti-rabbit secondary antibody (1:500, 60 min). After three more washing steps in TBST, samples were incubated at room temperature in VECTASTAIN ABC-Peroxidase Kit (1:100, 60 min) and stained using SIGMAFAST™ 3,3′diaminobenzidine tablets according to manufacturer’s instructions. Sections were washed twice in TBS, once in distilled water and counterstained with Hemalum solution acidic according to Mayer. Brain slices were air-dried overnight and microscope glasses were coverslipped with Entellan® (107960, Merck, Darmstadt, Germany). Comparable to rat staining, negative controls were processed in parallel by omitting the primary antibody. To determine the cellular expression profile of HSPA5 and HSPH4 in rat tissue, we performed double-immunolabeling using markers for principal neurons (NeuN), for astrocytes (GFAP) and for microglia (Iba1). Six animals, three from CTR and three from SE groups, were randomly selected among the entire pool of animals for each time point (http://randomizer.org). Sections were initially rinsed thrice in phosphate buffered saline containing 0.1% Tween-20 (PBST), followed by HIER as described above. Afterwards, a 60 min blocking step was 10

performed at room temperature using PBS added with 2% bovine serum albumin (A2153, Sigma-Aldrich, Germany) and 0.01% cold water fish skin gelatin (G7041, Sigma-Aldrich, Germany) in order to prevent unspecific bindings. Primary antibodies for the proteins of interest (rabbit anti-HSPA5 - 1:2500, rabbit anti-HSPH4 - 1:150) were incubated overnight (4°C, 16h) in combination with either antibodies for NeuN (1:500), GFAP (1:1000), or Iba1 (1:1000). Following three PBST washes, sections were incubated with secondary antibodies. Alexa Fluor 594-conjugated goat anti-rabbit antibody (1:1000, 60 min, 24°C) was used to label HSPA5, while biotinylated goat anti-rabbit (1:1000, 60 min, 24°C) together with Cy3conjugated streptavidin (1:2000, 60 min, 24°C) was used to label HSPH4. Alexa Fluor 488conjugated goat anti-mouse antibody was used to label both NeuN and GFAP (1:1000, 60 min, 24°C) while for Iba1 labelling the Cy2-conjugated donkey anti-goat antibody (1:1000, 90 min, 24°C) was chosen. Finally, after three further washing steps in PBS, slices were counterstained using Hoechst 33342, mounted on microscope glasses and coverslipped with Immunoselect Antifading Mounting Medium. HSPH4 and HSPA5 stereological cell count in rats Rat brain sections distanced from each other by 720 μm were processed for stereological cell counting. HSPA5 and HSPH4 positive cells were counted using the systematic random sampling (Gundersen & Jensen, 1987) in following areas: 

Hippocampus [bregma from -2.1 to -6.1]: hilus of the dentate gyrus, Cornu Ammonis region 1 (CA1) and region 3 (CA3)



Parahippocampal cortex: perirhinal cortex [bregma from -3.1 to -7.1], lateral and medial entorhinal cortex [bregma from -2.4 to -7.1], piriform cortex [bregma from 0.0 to -4.9] layer 1 (Pir1), layer 2 (Pir2) and layer 3 (Pir3) were analyzed separately given their different contribution to epileptogenesis (Loscher et al., 1995; Vaughan & Jackson, 2014)

Unbiased stereological cell counting was performed using an optical brightfield microscope (DMLB, Leica Microsystems, RRID: SCR_008960, Germany) equipped with Bio Point 2 Motorized Stage (Ludl Electronic Products, USA) and a color CCD camera (CX9000 - MBF Bioscience USA). The person performing the quantification was blinded and all samples were randomized before the analysis (http://randomizer.org). Samples were evaluated using Stereo Investigator software (version 11 - MBF Bioscience, USA) with area estimation performed at low magnification (5x/0.12 NA) and cell count performed at high magnification (40x/0.65 NA). For each region, calculated volumes were estimated by Cavalieri’s principle (Gundersen 11

& Jensen, 1987) and the total number of neurons was estimated using the optical fractionator counting method (West et al., 1991). The number of HSPH4 and HSPH5 positive cells, returned from the software as estimated population using Mean Section Thickness, were expressed as cell density using the calculated volume: 𝐶𝑒𝑙𝑙 𝐷𝑒𝑛𝑠𝑖𝑡𝑦 =

(𝐻𝑠𝑝𝐻4 𝑜𝑟 𝐻𝑠𝑝𝐴5) 𝑝𝑜𝑠𝑖𝑡𝑖𝑣𝑒 𝑐𝑒𝑙𝑙𝑠 𝐶𝑎𝑙𝑐𝑢𝑙𝑎𝑡𝑒𝑑 𝑉𝑜𝑙𝑢𝑚𝑒

The software returns the coefficient of error (CE), an index expressing the precision for each stereological cell count, in both m=0 and m=1; we opted for the m=1 since it has already been described (Gundersen et al., 1999) that biological samples are best described by this metric. In order to evaluate the study design, in addition to the estimate of within-animal variation by the CE, the variability between animals per region was calculated as a standard deviation (SD) and a coefficient of variation (CV = SD/mean), that measures the inter-animal variability. Both coefficients were afterwards used to evaluate study design by means of a CE2/CV2 ratio. [CE2/CV2] ≤ 0.5 was considered acceptable (Gundersen, 1986), as also reported by other studies (Rajkowska et al., 2016). All precision parameters for stereological cell counting are gathered and reported in Table S2 (for HSPA5) and Table S3 (for HSPH4). HSPH4 and HSPA5 optical density and area in dogs Since canine tissue samples were collected post mortem, a highly standardized tissue sampling procedure as the one carried out in rats was impossible. For this reason, a stereological cell count could not be performed and instead we quantified the level of HSPA5 and HSPH4 by means of optical density (OD) and area of the staining for these two proteins. Images were captured at 20x magnification using an Olympus BH2 microscope equipped with a single chip charge-coupled device (CCD) color camera (Axiocam; Zeiss, Göttingen, Germany) and an AMD Athlon 64 processor-based computer with image capturing software (AxioVision v4.6, RRID:SCR_002677, Zeiss, Germany). Images were analyzed by ImageJ software (ImageJ v1.51, RRID:SCR_003070, NIH) as described by Schneider et al. (2012). Regions of interest included hippocampal dentate gyrus, hilus, CA1, CA3 and parahippocampal cortex. Due to lack of standardized sources regarding detailed anatomy of canine brain, we decided, unlike in rats, not to divide the parahippocampal cortex into sub-

12

regions. In each region, five representative pictures were acquired at 40x magnification, resulting in up to 25 visual fields (297.22 x 222.70 µm) obtained from each animal. For OD analysis, at first we performed the calibration in accordance with instructions on the software website (ImageJ). Images were then analyzed as follows: initially pictures were cropped to contain only the region of interest and exclude artefacts. Afterwards, images were processed with the color deconvolution plugin (vector Hematoxylin-DAB) described by Ruifrok and Johnston (2001) that generates three 8-bit images (brown, blue and green), of which only the brown one, corresponding to DAB, was further processed. Histogram-based automatic triangle threshold (Zack et al., 1977) was applied to images in order to select specific signal over the background. Thresholded area and mean OD values from all pictures analyzed were averaged for each region and further used for statistical analysis. HSPH4 and HSPA5 colocalization imaging Confocal laser scanning microscopy was performed on a Carl Zeiss LSM-880 microscope equipped with argon (458, 488 and 514 nm), DPSS 561-10 (405 nm) and HeNe633 (633 nm) lasers (Zeiss LSM 800 with Airyscan, RRID: SCR_015963). Images were captured at resolution of 1024 X 1024 pixels and in order to improve the signal-to-noise ratio, two scans of one focal plane were averaged. Images were obtained using Plan-Apochromat 20x/0.8 NA and Plan-Apochromat 63x/1.4 NA objectives, then further processed with Zeiss built-in software

(ZEN

Digital

Imaging

for

Light

Microscopy

-

black

edition

v.2.3,

RRID:SCR_013672). All animals were visually inspected in hilus, CA1, CA3, perirhinal and piriform cortices to assess and locate the presence of HSPH4 and HSPA5 labelled cells. Where signal was located, 5 to 7 single confocal plane representative images were acquired at low magnification both in ipsilateral and contralateral (to the electrode) hemispheres. In case the probes signal was not clearly identifiable, both, higher magnification and z-stack micrographs were captured in order to increase image resolution. Statistics Statistical analysis was performed with SPSS v24 (SPSS, RRID:SCR_002865, USA) and Prism 5.04 (GraphPad Prism, RRID: SCR_002798; San Diego, USA). Ipsilateral and contralateral estimated cell counts from each time point in all regions, as well as the volumes calculated by the Stereo Investigator software were compared using the one-way ANOVA followed by Bonferroni post-hoc test. 13

In rats, HSPH4 and HSPA5 cell density between rats with SE and control rats as well as between different time points of the experiment were analyzed using two-way ANOVA followed by Bonferroni post-hoc test. In dogs, HSPH4 and HSPA5 OD between SE and control was tested in all regions within each time point by using the one-way ANOVA followed by Bonferroni post-hoc test. Data are indicated as mean ± SEM and we considered a p value < 0.05 statistically significant.

14

Results HSPH4 and HSPA5 expression in rats Immunohistochemistry for HSPA5 and HSPH4 provided information about their distribution across the investigated brain regions. At each time point, both, SE and control rats, presented a similar staining pattern, with high abundance in all investigated brain regions and signal distributed mostly in cell soma. Most of immunopositive cells seemed to resemble principal neurons, with an exception of several positively stained cells found in the layer 1 of the piriform cortex, which matched the morphology of glial cells, possibly astrocytes. No difference in staining intensity nor in localization were observed between the time points. HSPA5 was found to be in the very proximity of the cell nucleus, as expected from an ER chaperone, but also in cytosol reaching in some cases the processes. In contrast, HSPH4 had a more diffuse expression in the extracellular space in the regions surrounding the pyramidal layer, namely the stratum lacunosum-moleculare, stratum radiatum, stratum lucidum, and the stratum oriens. In the stereological analysis, no difference was demonstrated between the ipsilateral and contralateral side, neither in cell count nor in volume calculation. For that reason, results from both sides were summed up for a respective region in each animal and expressed as a cell density. SE modified the expression of these proteins differently at the three time points with a stronger effect in the latency phase, a marginal effect in the acute, and no effect in the chronic phase. Stereological cell counting revealed that HSPA5 was up-regulated in the parahippocampal cortex (Fig 1D-H). Specifically, in piriform layer 1 of animals experiencing SE, we found in the early post-insult and latency phase a 3.5-fold and a 3.6-fold increase in cell density, respectively (F [1, 30] = 19.844, p = 0.0001). Cells in this region presented also a darker staining in SE animals (Fig 1J, L) compared to age matched controls (Fig 1I, K). In the other parahippocampal cortex regions, SE did not affect the number of cells in all time points. Quantification of HSPA5 levels revealed a tangible effect of SE on this protein both in the hippocampus (Fig 1A-C) and in the parahippocampal cortex. In the hippocampal area, a downregulation of HSPA5 levels in animals with a SE history has been observed in all analyzed subregions (hilus: F [1, 30] = 7.543, p = 0.01; CA3: F [1, 30] = 4.633, p = 0.04; CA1: F [1, 30] = 5.727, p = 0.023). The analysis of the influence of time course on the expression of HSPA5 revealed statistically significant changes in all examined regions except in CA1 and layer 1 of the piriform cortex (hilus: F [2, 30] = 3.368, p = 0.048; CA3: F [2, 30] = 6.197, p = 0.006; entorhinal cortex: F [2, 30] = 6.116, p = 0.006; perirhinal cortex: F [2, 30] = 3.181, p = 0.004; 15

piriform cortex layer 2: F [2, 30] = 3.686, p = 0.037; piriform cortex layer 3: F [2, 30] = 6.404, p = 0.005). Post-hoc analysis revealed a significant increase of HSPA5 levels in the chronic phase as compared to the latency phase (hilus: p = 0.046; CA3: p = 0.005; entorhinal cortex: p = 0.013; perirhinal cortex: p = 0.005; piriform cortex layer 2: p = 0.034; piriform cortex layer 3: p = 0.004) as well as in comparison to the acute post-insult phase (entorhinal cortex: p = 0.017; perirhinal cortex: p = 0.035). Quantification of HSPH4 revealed a less discernible effect of SE on this protein both in the hippocampal area (Fig 2A-C) and in the parahippocampal cortex (Fig 2D-H). Similarly to HSPA5, a down-regulation of HSPH4 has been observed in the hilus of the dentate gyrus (F [1, 29] = 10.117, p = 0.003). In the perirhinal cortex we were able to elucidate a trend for an an increase of HSPH4 expression (F [1, 30] = 4.108, p = 0.052). Time-related changes were observed in layer 3 of the piriform cortex (F [2, 30] = 3.717, p = 0.036). Post-hoc testing revealed a significant up-regulation of HSPH4 expression in the chronic phase as compared to the latency phase (p = 0.038). HSPH4 staining in animals with SE (Fig 2M-P) appeared darker than respective controls (Fig 2I-L), both in the cell soma and in the extracellular space. HSPH4 and HSPA5 expression in dogs Analysis of immunohistochemically stained brain tissue revealed a strong cytoplasmatic signal of HSPA5 present in all regions of interest. In the hippocampus, the majority of positive cells were present in CA1 (Fig 3E) and CA3 pyramidal layer as well as in the polymorphic cell layer of dentate gyrus (hilus). In the parahippocampal cortex, the most cell dense region was the layer 2 of piriform cortex. Similarly to rats, the morphology and location of HSPA5 stained cells resembled pyramidal neurons and, with minor extent, astroglia. Signal from HSPA5 was detected predominantly in the cell soma. No group difference became evident when analyzing protein expression in the different groups based on stained area (Fig 3A-F) or OD (data not shown). HSPH4 analysis revealed a strong cytoplasmic signal in all regions of interest, with positive cells located in CA1 (Fig 3H) and CA3 pyramidal layer as well as hilus of the hippocampus. In the latency phase, the majority of stained cells were present in layer 1 and 2 of the piriform cortex. Cell morphology and localization was indicative of pyramidal neurons with a predominant signal in the soma. In several cells the signal proved to be distributed throughout the cytoplasm with staining also visible in processes. In addition, a diffuse staining was evident in the extracellular space.

16

Quantitative analysis for this protein revealed a significant increase of stained area in CA1 (F [3, 43] = 5.569, p = 0.0045; Fig 3K) in animals with structural epilepsy compared to the patient control group. In this region, post-hoc comparison between individual groups revealed an enlargement in the labelled area in animals with structural epilepsy as compared to the patient control group of 151%. In the molecular layer of the dentate gyrus the HSPH4 immunopositive area tended to be increased as a consequence of structural epilepsy (vs. patient control group: F [3, 42] = 2.495, p = 0.0729). HSPA5 and HSPH4 cell profiling In rat hippocampus and parahippocampal cortex sub-regions, we analyzed brain cell types expressing both HSPA5 and HSPH4 by means of immunofluorescent double labeling and laser confocal imaging. In order to do that, we used markers specific for neurons (NeuN), microglia cells (Iba1) and astrocytes (GFAP). Immunostaining for HSPA5 (Fig. 4) confirmed that this protein is expressed in neurons in both experimental groups. In all the regions investigated, NeuN signal (Fig. 4B) colocalized with HSPA5 signal (Fig. 4A) with the latter being distributed around the cell nucleus counterstained with Hoechst 33342 (Fig 4C). In contrast, Iba1 signal (Fig. 4E) did not colocalize with HSPA5 signal (Fig. 4D), with morphologies of stained cells differing significantly as visible in the merge image counterstained with Hoechst 33342 (Fig 4F). Double immunostaining of HSPA5 (Fig. 4G) with the astrocytic marker GFAP (Fig. 4H) appeared restricted to neuronal cells in almost all brain regions analyzed. An exception was visible in piriform cortex layer 1 where few cells exhibited colocalization between the two markers (Fig. 4I). This colocalization has been validated by a high magnification Z-stack image (Fig. 5). Immunostaining for HSPH4 (Fig. 6) also confirmed the expression of this protein in neurons in both experimental and control groups. In all regions of interest, NeuN signal (Fig. 6B) colocalized with HSPH4 signal (Fig. 6A), with the latter presenting a cytoplasmic localization as evident in the merge image counterstained with Hoechst 33342 (Fig 6C). Iba1 signal (Fig. 6E) did not colocalize with HSPH4 signal (Fig. 6D), with morphologies of immunopositive cells differing significantly as visible in the Hoechst 33342 counterstained image (Fig 6F). Likewise, the labeling of HSPH4 (Fig. 6G) with GFAP (Fig. 6H) did not demonstrate any colocalization of these two proteins (Fig. 6I). The colocalization patterns looked similar in both experimental and control groups, with no temporal nor spatial influence on the quality of the signal, with an exception of HSPA5-GFAP staining, which showed double-immunolabelled cells in the piriform cortex of SE animals, but 17

not in the controls. For this reason, as well as to avoid the redundancy of the given information, one representative overview image is presented for each of the stainings. To underline the difference present in HSPA5 colocalization with GFAP, one high magnification Z-stack image is presented.

18

Discussion In the present study, the expression pattern of two endoplasmic heat shock proteins, HSPA5 and HSPH4, was analyzed during the course of epileptogenesis in adult rats. Comprehensive information about temporal and spatial epileptogenesis-associated alterations in expression patterns of the two proteins are crucial for a better understanding of the role of ER-stress related processes in the development of TLE. Moreover, respective data provides valuable information about their role as possible target candidates for the development of therapeutic approaches modulating the unfolded protein response. The time course of expression patterns informs additionally about time-windows for a putative anti-epileptogenic intervention. Based on a large-scale proteomic analysis, we have previously obtained first evidence for a regulation of HSPA5 and HSPH4 in different phases of epileptogenesis in the electrical rat SE model (Walker et al., 2016). In the present immunohistochemistry study, we confirmed expression alterations of HSPA5 and HSPH4. In addition, our data sets provide detailed information about the expression patterns in different sub-regions of the hippocampus and the parahippocampal cortex. Analysis of both proteins of interest demonstrated changes in expression patterns following SE with a significant up- or down-regulation in sub-regions of the temporal lobe crucial for the development of TLE. Moreover, analysis in most regions indicated a significant increase of HSPA5 levels in the chronic phase in comparison to the latent phase. The latent phase is defined as the time after the brain insult when molecular and cellular mechanisms leading to the onset of clinical seizures are progressing (Pitkanen et al., 2015). Our findings reveal that the most pronounced regulation of the two ER chaperones occurs subsequently to this phase, when abnormal neuronal networks are already established. This finding is of particular interest, since most of the studies describe a decrease of HSPA5 levels with age (Erickson et al., 2006; Brown & Naidoo, 2012; Salganik et al., 2015). One has to however account for the fact that animals used in above-mentioned studies are usually older (about 24 months) than the ones used in our study (3-5 months). Additionally, a study performed in female rats (as used in our experiments) did not indicate any changes in HSPA5 levels with age (Gleixner et al., 2014). In earlier studies, repeated seizure induction was shown to trigger up-regulation of markers of ER stress, including HSPA5, in kindling models in rats (Chihara et al., 2011) and mice (Zhu et al., 2017). Considering that repeated kindled seizures result in an increased seizure susceptibility and reduced seizure thresholds, the findings along with our data from an epileptogenesis model may indicate that ER stress represents a hallmark of processes resulting in the generation of 19

hyperexcitable networks. However, an influence of the electrode also needs to be considered. Implantation of neural probes has been described as a factor triggering transcription of genes related to endoplasmic reticulum stress (Ereifej et al., 2018). No data regarding HSPA5 and HSPH4 expression as a follow-up of electrode implantation is available so far, but it is not unlikely that the increase in the proteins’ levels with time has at least been partly triggered by the surgery itself. Nevertheless, our results demonstrate significant changes in HSPA5 and HSPH4 levels in stimulated animals as compared to sham-operated controls supporting the hypothesis that status epilepticus and epileptogenetic processes influence the expression of the two proteins. The expression pattern of both proteins varied in response to the epileptogenic insult, but this variation was restricted regionally. The areas with the most pronounced expression alterations included the piriform cortex as well as the hippocampus. In the piriform cortex, we detected changes in expression levels of HSPA5 in layer 1, the most superficial part of the region consisting mostly of axons and glial cells. The piriform cortex is rich in projections to and from the entorhinal and the perirhinal cortex and the amygdala, and it demonstrates low inhibitory GABAergic connectivity, especially in its rostral portion, called area tempestas (Vismer et al., 2015). Its unique circuitry is a foundation of its role as a region susceptible to convulsants and able to generate seizures (Piredda & Gale, 1985). Considering the contribution of the piriform cortex to the development of hyperexcitability during epileptogenesis, the regulation of ER chaperones in this brain region might indicate an involvement in the pathophysiological mechanisms of epilepsy development. On the other hand, induction of HSPA5 in the piriform cortex, may just reflect its particular susceptibility to neuronal damage following an epileptogenic insult (Vismer et al., 2015). In this context, it is of interest that ER stress is discussed as a relevant contributor to neuronal damage (Scheper & Hoozemans, 2015) and that the ER chaperone can serve a protective function (Kitao et al., 2001b; Kitao et al., 2004; Casas, 2017; Louessard et al., 2017). Levels of both proteins of interest were demonstrated to decrease in the hippocampus. This outcome might seem counterintuitive at the first sight, since one would assume high excitotoxicity in the brain following status epilepticus would evoke a strong stress response from ER. However, it is possible that ischemic conditions associated with status epilepticus (Fabene et al., 2007) lead to an impairment of ER function and therefore an ER stress reaction cannot be fully evoked (Paschen & Mengesdorf, 2005). This in consequence could lead to neuronal cell death when the protective mechanisms of UPR are not sufficient. On the other hand, even a single but excessive induction of ER stress responses following status epilepticus 20

could lead to detrimental effects via the apoptotic pathway (Han et al., 2013). Inhibition of ER stress responses has already been described as protective against excitotoxic neuronal injury following seizures (Sokka et al., 2007b). However, more studies are needed, especially at early time points, to determine exactly which events occur in the hippocampus following SE. Nevertheless, changes of expression of ER chaperones in hippocampal subregions further support a conclusion that ER stress might be functionally related to the consequences of an epileptogenic insult. Qualitative analysis of HSPA5 and HSPH4 expression in different brain cell populations revealed a predominant colocalization of both proteins with the neuronal marker NeuN and no colocalization with the microglial marker Iba1. Restricted colocalization has been observed between the astrocytic marker GFAP and HSPA5, but not HSPH4. These findings suggest that ER stress during epileptogenesis is rather limited to neurons, which are the main cells involved in development of dysfunctional circuitry in the course of epileptogenesis. The glial response to the conditions following SE was in this case either too weak, or a transient regulation was missed with the three selected time points. The half-life of HSPH4 is estimated to be around 10.77 hours (Xiao & Wu, 2017), so in case of a very rapid and transient glial response it might have been too late to notice alterations in glia. However, the half-life of HSPA5 is relatively long, reaching from over 24 hours in cell culture to over 3 days in tissues so that it is rather unlikely that we missed an early induction in glial cells (Preissler & Ron, 2018). Neuronal HSPA5 and HSPH4 can exert protective effects as a response to stress factors. In a stroke model in mice HSPA5 protected neurons by binding to and decreasing an overactivation of the protein kinase RNA-like endoplasmic reticulum kinase (PERK) arm of the UPR which leads to deleterious consequences for cell homeostasis (Louessard et al., 2017). The protein plays also a relevant role in neuroprotection in many common neurodegenerative diseases, like Alzheimer’s disease, Parkinson’s disease and amyotrophic lateral sclerosis, promoting cell survival and decreasing degeneration (Casas, 2017). Neurons have been demonstrated to express HSPH4 in response to glutamate and to show a higher viability following kainate administration in transgene mice overexpressing the protein (Kitao et al., 2001b). Protein overexpression was also demonstrated following hypoxic stress in cultured neurons (Tamatani et al., 2001). Since HSPA5 and HSPH4 seem to be of high importance for promotion of neuronal survival and decrease of neurodegeneration, both in physiological and pathological conditions, it is plausible that similar neuroprotective mechanisms might apply in case of the early consequences of a status epilepticus as well as subsequent epileptogenic processes.

21

In addition to regulation of protein expression, one has to consider posttranslational regulation of function as well. In the ER, HSPA5 is present in two states, active or inactive, depending on AMP-binding, which locks the protein in low-affinity state for its substrates (Preissler & Ron, 2018). The chaperone may also be deposited into a pool of temporarily inactive proteins in a process of oligomerization, in which single HSPA5 polypeptides bind to each other’s substrate binding locus in a competitive manner (Preissler & Ron, 2018). Therefore, in an environment with lacking chaperone substrates, HSPA5 can still be abundant in cells, but either in its inactive AMPylated form, or deposited in oligomers (Preissler & Ron, 2018). While posttranslational alterations of HSPA5 were not in the scope of this study, it might be of future interest to investigate an impact of epileptogenic insults on HSPA5 cellular fractions. The unfolded protein response is a long-discussed target of many putative therapeutic intervention approaches. It has been described as a possible therapeutic object in anticancer therapy (Wang et al., 2018), in cardiovascular disease (Zhang et al., 2019) as well as in neurodegenerative diseases (Hughes & Mallucci, 2019; Martinez et al., 2019). Modulation of HSPA5 levels specifically has been proposed in cancer therapy, in which cleavage of HSPA5 polypeptides leads to a lower cancer cell survival (Backer et al., 2011). Several therapeutic agents regulating HSPA5 and exerting neuroprotective effects have been described in models of Parkinson’s disease (Enogieru et al., 2019). The function of the endoplasmic reticulum might be a possible therapeutic target also considering the ER-mitochondria cross-talk and a functional link of ER stress to mitochondria. There is a close contact between the two organelles, with 20% of the mitochondrial surface linked directly to the ER (Kornmann et al., 2009) and their interaction is critical for proper calcium signaling flowing from the ER into the mitochondria (Malhotra & Kaufman, 2011). Sigma-1R receptor present at these binding sites is directly functionally connected to HSPA5, and during the stress conditions its activation has been demonstrated to perpetuate calcium signaling and ultimately to support cell survival (Malhotra & Kaufman, 2011; Nguyen et al., 2015). Targeting sigma-1R is a concept already being discussed in epileptology, since allosteric modulation of the receptor provided anti-seizure effects in status epilepticus rat models (Guo et al., 2015; Vavers et al., 2017). Additional modulation of HSPA5, which plays a pivotal role in sigma-1R functionality might be of further advantage in developing therapy approaches targeted towards this complex. The rat model of electrical SE takes into consideration highly standardized, laboratory conditions. On the other hand, analysis performed in canine brain tissue allowed us to better understand changes that HSPA5 and HSPH4 undergo in the more unpredictable clinical 22

environment. Data from patients handled in the clinics can be directly compared to the data from the chronic phase with spontaneous recurrent seizures in the rat model, since this phase mirrors the actual presentation of TLE. In rats, during this phase, we demonstrated an upregulation in comparison to the latency and the early post-insult phase in the cases of both HSPA5 and HSPH4. However, to have a better insight into possible discrepancies between the clinical and experimental conditions, we decided to include two groups for comparison: a patient control group, comprising animals of different breed, age and history of treatment, which mimicked clinical conditions, and an experimental control group with dogs of the same breed, age and clinical history, kept under standardized conditions and euthanized with subsequent immediate dissection. Not all dogs from the patient control group were euthanized. Some of the animals experienced a final disease state and agony phase due to a natural non-neurological disorder, which regardless of the underlying disorder can cause hypoxia and ischemia affecting brain cells and triggering central responses of heat shock chaperone proteins (Wang et al., 1993). However, the lack of statistically significant differences in HSPA5 and HSPH4 levels between the patient and the experimental groups suggests that such changes are not relevant for expression of these two proteins. Analysis performed in the immunohistochemically stained canine tissue revealed no differences between the groups in HSPA5, however an increase of the stained area was observed in CA1 region in the dogs diagnosed with structural epilepsy in the case of HPSH4. The nature of this discrepancy between the rat and the dog models remains unclear and might be related to species differences and influenced by drug treatment, duration of the disease, age, breed, co-morbidities and others. However, this finding might further advocate that ER stressrelated response may be involved also in the chronic phase of the disease, so therapeutic targeting of the respective heat shock protein could be advantageous when the symptoms are fully developed. Further research involving more clinical cases is nevertheless necessary to elucidate the relevance of the two proteins to TLE manifestation. Some evidence links the UPR to inflammatory processes, particularly to the regulation of Tolllike receptors expression (Timberlake et al., 2018). A recent study from our working group described epileptogenesis-related alterations in expression of another heat shock protein, HSPA1A, a well-known ligand of Toll-like receptor 4. HSPA1A was clearly up-regulated in SE animals in the early post-insult phase with some temporal lobe regions demonstrating an up-regulation also in the latency phase (Gualtieri et al., 2019). This further connects HSPs and processes involving them to epileptogenesis, which makes them promising candidates for 23

future therapeutic targets. In the case of HSPA5 and HSPH4, however, additional studies with genetic and pharmacological modulation of both proteins with subsequent analysis of the consequences in animal models of TLE are necessary to elucidate their impact on the process of the development of the disease. In conclusion, we provided an extensive characterization of alterations following status epilepticus in the rat model as well as present in the developed form of epilepsy in clinical canine cases. HSPA5 was up-regulated in a temporally and spatially restricted manner. The most prominent up-regulation was observed in the piriform cortex and during the chronic phase. In contrast, both, HSPA5 and HSPH4 exhibited reduced expression levels in the hippocampus as a consequence of the status epilepticus history. This information could be of importance while developing therapeutic strategies targeting these two proteins.

24

Acknowledgements Research in H. Potschka’s group has been supported by grants of the Deutsche Forschungsgemeinschaft (DFG PO 681/5-2 and PO 681/8-1). The authors thank Sarah Driebusch, Sieglinde Fischlein, Luzie Rettenbeck, Sabine Sass, Isabel Seiffert, Claudia Siegl and Andreas Walker for their excellent technical assistance.

Conflict of Interest Disclosure The authors declare that they do not have any conflict of interest. 

25

References Aksoy, M.O., Kim, V., Cornwell, W.D., Rogers, T.J., Kosmider, B., Bahmed, K., Barrero, C., Merali, S., Shetty, N. & Kelsen, S.G. (2017) Secretion of the endoplasmic reticulum stress protein, GRP78, into the BALF is increased in cigarette smokers. Respir Res, 18, 78. Backer, M.V., Backer, J.M. & Chinnaiyan, P. (2011) Targeting the unfolded protein response in cancer therapy. Methods Enzymol, 491, 37-56. Behnke, J., Feige, M.J. & Hendershot, L.M. (2015) BiP and its nucleotide exchange factors Grp170 and Sil1: mechanisms of action and biological functions. Journal of molecular biology, 427, 1589-1608. Benarroch, E.E. (2011) Heat shock proteins: multiple neuroprotective functions and implications for neurologic disease. Neurology, 76, 660-667. Berendt, M., Farquhar, R.G., Mandigers, P.J., Pakozdy, A., Bhatti, S.F., De Risio, L., Fischer, A., Long, S., Matiasek, K., Munana, K., Patterson, E.E., Penderis, J., Platt, S., Podell, M., Potschka, H., Pumarola, M.B., Rusbridge, C., Stein, V.M., Tipold, A. & Volk, H.A. (2015) International veterinary epilepsy task force consensus report on epilepsy definition, classification and terminology in companion animals. BMC Vet Res, 11, 182. Braakman, I. & Bulleid, N.J. (2011) Protein folding and modification in the mammalian endoplasmic reticulum. Annual review of biochemistry, 80, 71-99. Brandt, C., Glien, M., Potschka, H., Volk, H. & Loscher, W. (2003a) Epileptogenesis and neuropathology after different types of status epilepticus induced by prolonged electrical stimulation of the basolateral amygdala in rats. Epilepsy Res, 55, 83-103. Brandt, C., Glien, M., Potschka, H., Volk, H. & Löscher, W.J.E.r. (2003b) Epileptogenesis and neuropathology after different types of status epilepticus induced by prolonged electrical stimulation of the basolateral amygdala in rats. 55, 83-103. Brown, M.K. & Naidoo, N. (2012) The endoplasmic reticulum stress response in aging and age-related diseases. Front Physiol, 3, 263. Casas, C. (2017) GRP78 at the Centre of the Stage in Cancer and Neuroprotection. Front Neurosci, 11, 177. Chen, J., Guo, H., Zheng, G. & Shi, Z.-N. (2013) Region-specific vulnerability to endoplasmic reticulum stress-induced neuronal death in rat brain after status epilepticus. J Biosciences, 38, 877-886. 26

Chen, J., Zheng, G., Guo, H. & Shi, Z.-n. (2014) Role of endoplasmic reticulum stress via the PERK signaling pathway in brain injury from status epilepticus. Journal of Molecular Neuroscience, 53, 677-683. Chihara, Y., Ueda, Y., Doi, T. & Willmore, L.J. (2011) Role of endoplasmic reticulum stress in the amygdaloid kindling model of rats. Neurochem Res, 36, 1834-1839. Delpino, A. & Castelli, M. (2002) The 78 kDa glucose-regulated protein (GRP78/BIP) is expressed on the cell membrane, is released into cell culture medium and is also present in human peripheral circulation. Biosci Rep, 22, 407-420. Dragovic, Z., Broadley, S.A., Shomura, Y., Bracher, A. & Hartl, F.U. (2006) Molecular chaperones of the Hsp110 family act as nucleotide exchange factors of Hsp70s. The EMBO journal, 25, 2519-2528. Easton, D.P., Kaneko, Y. & Subjeck, J.R. (2000) The Hsp110 and Grp170 stress proteins: newly recognized relatives of the Hsp70s. Cell Stress Chaperon, 5, 276. Enogieru, A.B., Omoruyi, S.I., Hiss, D.C. & Ekpo, O.E. (2019) GRP78/BIP/HSPA5 as a Therapeutic Target in Models of Parkinson's Disease: A Mini Review. Adv Pharmacol Sci, 2019, 2706783. Ereifej, E.S., Rial, G.M., Hermann, J.K., Smith, C.S., Meade, S.M., Rayyan, J.M., Chen, K., Feng, H. & Capadona, J.R. (2018) Implantation of Neural Probes in the Brain Elicits Oxidative Stress. Front Bioeng Biotechnol, 6, 9. Erickson, R.R., Dunning, L.M. & Holtzman, J.L. (2006) The effect of aging on the chaperone concentrations in the hepatic, endoplasmic reticulum of male rats: the possible role of protein misfolding due to the loss of chaperones in the decline in physiological function seen with age. J Gerontol A Biol Sci Med Sci, 61, 435-443. Fabene, P.F., Merigo, F., Galie, M., Benati, D., Bernardi, P., Farace, P., Nicolato, E., Marzola, P. & Sbarbati, A. (2007) Pilocarpine-induced status epilepticus in rats involves ischemic and excitotoxic mechanisms. PLoS One, 2, e1105. Gleixner, A.M., Pulugulla, S.H., Pant, D.B., Posimo, J.M., Crum, T.S. & Leak, R.K. (2014) Impact of aging on heat shock protein expression in the substantia nigra and striatum of the female rat. Cell Tissue Res, 357, 43-54. Gualtieri, F., Nowakowska, M., von Ruden, E.L., Seiffert, I. & Potschka, H. (2019) Epileptogenesis-Associated Alterations of Heat Shock Protein 70 in a Rat Post-Status Epilepticus Model. Neuroscience.

27

Gundersen, H.J. (1986) Stereology of arbitrary particles. A review of unbiased number and size estimators and the presentation of some new ones, in memory of William R. Thompson. J Microsc, 143, 3-45. Gundersen, H.J. & Jensen, E.B. (1987) The efficiency of systematic sampling in stereology and its prediction. J Microsc, 147, 229-263. Gundersen, H.J.G., Jensen, E.B.V., Kieu, K. & Nielsen, J.J.J.o.m. (1999) The efficiency of systematic sampling in stereology—reconsidered. 193, 199-211. Guo, L., Chen, Y., Zhao, R., Wang, G., Friedman, E., Zhang, A. & Zhen, X. (2015) Allosteric modulation of sigma-1 receptors elicits anti-seizure activities. Br J Pharmacol, 172, 4052-4065. Han, F., Yan, S. & Shi, Y. (2013) Single-prolonged stress induces endoplasmic reticulumdependent apoptosis in the hippocampus in a rat model of post-traumatic stress disorder. PLoS One, 8, e69340. Hughes, D. & Mallucci, G.R. (2019) The unfolded protein response in neurodegenerative disorders - therapeutic modulation of the PERK pathway. FEBS J, 286, 342-355. ImageJ, N.I.o.H., Bethesda, Optical Density Calibration. Jaronen, M., Goldsteins, G. & Koistinaho, J. (2014) ER stress and unfolded protein response in amyotrophic lateral sclerosis—a controversial role of protein disulphide isomerase. Frontiers in cellular neuroscience, 8, 402. Keck, M., Androsova, G., Gualtieri, F., Walker, A., von Ruden, E.L., Russmann, V., Deeg, C.A., Hauck, S.M., Krause, R. & Potschka, H. (2017) A systems level analysis of epileptogenesis-associated proteome alterations. Neurobiology of disease, 105, 164178. Keck, M., van Dijk, R.M., Deeg, C.A., Kistler, K., Walker, A., von Rüden, E.-L., Russmann, V., Hauck, S.M. & Potschka, H.J.N.o.d. (2018) Proteomic profiling of epileptogenesis in a rat model: Focus on cell stress, extracellular matrix and angiogenesis. 112, 119135. Kitao, Y., Hashimoto, K., Matsuyama, T., Iso, H., Tamatani, T., Hori, O., Stern, D.M., Kano, M., Ozawa, K. & Ogawa, S. (2004) ORP150/HSP12A regulates Purkinje cell survival: a role for endoplasmic reticulum stress in cerebellar development. J Neurosci, 24, 1486-1496. Kitao, Y., Ozawa, K., Miyazaki, M., Tamatani, M., Kobayashi, T., Yanagi, H., Okabe, M., Ikawa, M., Yamashima, T. & Stern, D.M. (2001a) Expression of the endoplasmic 28

reticulum molecular chaperone (ORP150) rescues hippocampal neurons from glutamate toxicity. The Journal of clinical investigation, 108, 1439-1450. Kitao, Y., Ozawa, K., Miyazaki, M., Tamatani, M., Kobayashi, T., Yanagi, H., Okabe, M., Ikawa, M., Yamashima, T., Stern, D.M., Hori, O. & Ogawa, S. (2001b) Expression of the endoplasmic reticulum molecular chaperone (ORP150) rescues hippocampal neurons from glutamate toxicity. J Clin Invest, 108, 1439-1450. Ko, A.R., Kim, J.Y., Hyun, H.W. & Kim, J.E. (2015) Endoplasmic reticulum (ER) stress protein responses in relation to spatio-temporal dynamics of astroglial responses to status epilepticus in rats. Neuroscience, 307, 199-214. Kornmann, B., Currie, E., Collins, S.R., Schuldiner, M., Nunnari, J., Weissman, J.S. & Walter, P. (2009) An ER-mitochondria tethering complex revealed by a synthetic biology screen. Science, 325, 477-481. Kozutsumi, Y., Segal, M., Normington, K., Gething, M.-J. & Sambrook, J. (1988) The presence of malfolded proteins in the endoplasmic reticulum signals the induction of glucose-regulated proteins. Nature, 332, 462. Loscher, W., Ebert, U., Wahnschaffe, U. & Rundfeldt, C. (1995) Susceptibility of different cell layers of the anterior and posterior part of the piriform cortex to electrical stimulation and kindling: comparison with the basolateral amygdala and "area tempestas". Neuroscience, 66, 265-276. Louessard, M., Bardou, I., Lemarchand, E., Thiebaut, A.M., Parcq, J., Leprince, J., Terrisse, A., Carraro, V., Fafournoux, P., Bruhat, A., Orset, C., Vivien, D., Ali, C. & Roussel, B.D. (2017) Activation of cell surface GRP78 decreases endoplasmic reticulum stress and neuronal death. Cell Death Differ, 24, 1518-1529. Maguire, J.J.E.c. (2016) Epileptogenesis: More than just the latent period. 16, 31-33. Malhotra, J.D. & Kaufman, R.J. (2011) ER stress and its functional link to mitochondria: role in cell survival and death. Cold Spring Harb Perspect Biol, 3, a004424. Marin-Briggiler, C.I., Gonzalez-Echeverria, M.F., Munuce, M.J., Ghersevich, S., Caille, A.M., Hellman, U., Corrigall, V.M. & Vazquez-Levin, M.H. (2010) Glucoseregulated protein 78 (Grp78/BiP) is secreted by human oviduct epithelial cells and the recombinant protein modulates sperm-zona pellucida binding. Fertil Steril, 93, 15741584. Martinez, A., Lopez, N., Gonzalez, C. & Hetz, C. (2019) Targeting of the unfolded protein response (UPR) as therapy for Parkinson's disease. Biol Cell, 111, 161-168.

29

Nguyen, L., Lucke-Wold, B.P., Mookerjee, S.A., Cavendish, J.Z., Robson, M.J., Scandinaro, A.L. & Matsumoto, R.R. (2015) Role of sigma-1 receptors in neurodegenerative diseases. J Pharmacol Sci, 127, 17-29. Ni, M. & Lee, A.S. (2007) ER chaperones in mammalian development and human diseases. FEBS letters, 581, 3641-3651. Ni, M., Zhang, Y. & Lee, A.S. (2011) Beyond the endoplasmic reticulum: atypical GRP78 in cell viability, signalling and therapeutic targeting. Biochem J, 434, 181-188. Ongerth, T., Russmann, V., Fischborn, S., Boes, K., Siegl, C. & Potschka, H. (2014) Targeting of microglial KCa3.1 channels by TRAM-34 exacerbates hippocampal neurodegeneration and does not affect ictogenesis and epileptogenesis in chronic temporal lobe epilepsy models. Eur J Pharmacol, 740, 72-80. Paschen, W. & Mengesdorf, T. (2005) Endoplasmic reticulum stress response and neurodegeneration. Cell Calcium, 38, 409-415. Paxinos, G. & Watson, C. (1998) The rat brain in stereotaxic coordinates. Academic Press, San Diego. Piredda, S. & Gale, K. (1985) A crucial epileptogenic site in the deep prepiriform cortex. Nature, 317, 623-625. Pitkanen, A., Lukasiuk, K., Dudek, F.E. & Staley, K.J. (2015) Epileptogenesis. Cold Spring Harb Perspect Med, 5. Potschka, H., Fischer, A., von Rüden, E.-L., Hülsmeyer, V. & Baumgärtner, W. (2013) Canine epilepsy as a translational model? Epilepsia, 54, 571-579. Preissler, S. & Ron, D. (2018) Early Events in the Endoplasmic Reticulum Unfolded Protein Response. Cold Spring Harb Perspect Biol. Rachel, A., Tyson, J.R. & Stirling, C.J. (1997) A novel subfamily of Hsp70s in the endoplasmic reticulum. Trends in cell biology, 7, 277-282. Racine, R.J. (1972) Modification of seizure activity by electrical stimulation. II. Motor seizure. Electroencephalogr Clin Neurophysiol, 32, 281-294. Rajkowska, G., Clarke, G., Mahajan, G., Licht, C., de Werd, H.V., Yuan, P., Stockmeier, C., Manji, H. & Uylings, H. (2016) Differential Effect of Lithium on Cell Number in Hippocampus and Prefrontal Cortex in Adult Mice: A Stereological Study. Biological psychiatry, 79, 204s-204s. 30

Raviol, H., Sadlish, H., Rodriguez, F., Mayer, M.P. & Bukau, B. (2006) Chaperone network in the yeast cytosol: Hsp110 is revealed as an Hsp70 nucleotide exchange factor. The EMBO journal, 25, 2510-2518. Rüden, E.L.v., Avemary, J., Zellinger, C., Algermissen, D., Bock, P., Beineke, A., Baumgärtner, W., Stein, V.M., Tipold, A. & Potschka, H. (2012) Distemper virus encephalitis exerts detrimental effects on hippocampal neurogenesis. Neuropathology and Applied Neurobiology, 38, 426-442. Ruifrok, A.C. & Johnston, D.A. (2001) Quantification of histochemical staining by color deconvolution. Anal Quant Cytol Histol, 23, 291-299. Salganik, M., Sergeyev, V.G., Shinde, V., Meyers, C.A., Gorbatyuk, M.S., Lin, J.H., Zolotukhin, S. & Gorbatyuk, O.S. (2015) The loss of glucose-regulated protein 78 (GRP78) during normal aging or from siRNA knockdown augments human alphasynuclein (alpha-syn) toxicity to rat nigral neurons. Neurobiol Aging, 36, 2213-2223. Scheper, W. & Hoozemans, J.J. (2015) The unfolded protein response in neurodegenerative diseases: a neuropathological perspective. Acta Neuropathol, 130, 315-331. Schneider, C.A., Rasband, W.S. & Eliceiri, K.W.J.N.m. (2012) NIH Image to ImageJ: 25 years of image analysis. 9, 671. Shaner, L. & Morano, K.A. (2007) All in the family: atypical Hsp70 chaperones are conserved modulators of Hsp70 activity. Cell Stress Chaperon, 12, 1. Sokka, A.-L., Putkonen, N., Mudo, G., Pryazhnikov, E., Reijonen, S., Khiroug, L., Belluardo, N., Lindholm, D. & Korhonen, L. (2007a) Endoplasmic reticulum stress inhibition protects against excitotoxic neuronal injury in the rat brain. Journal of Neuroscience, 27, 901-908. Sokka, A.L., Putkonen, N., Mudo, G., Pryazhnikov, E., Reijonen, S., Khiroug, L., Belluardo, N., Lindholm, D. & Korhonen, L. (2007b) Endoplasmic reticulum stress inhibition protects against excitotoxic neuronal injury in the rat brain. J Neurosci, 27, 901-908. Soti, C., Nagy, E., Giricz, Z., Vigh, L., Csermely, P. & Ferdinandy, P. (2005) Heat shock proteins as emerging therapeutic targets. Br J Pharmacol, 146, 769-780. Stetler, R.A., Gan, Y., Zhang, W., Liou, A.K., Gao, Y., Cao, G. & Chen, J. (2010) Heat shock proteins: cellular and molecular mechanisms in the central nervous system. Prog Neurobiol, 92, 184-211.

31

Su, Y. & Li, F. (2016) Endoplasmic reticulum stress in brain ischemia. International Journal of Neuroscience, 126, 681-691. Tamatani, M., Matsuyama, T., Yamaguchi, A., Mitsuda, N., Tsukamoto, Y., Taniguchi, M., Che, Y.H., Ozawa, K., Hori, O., Nishimura, H., Yamashita, A., Okabe, M., Yanagi, H., Stern, D.M., Ogawa, S. & Tohyama, M. (2001) ORP150 protects against hypoxia/ischemia-induced neuronal death. Nat Med, 7, 317-323. Timberlake, M., 2nd, Prall, K., Roy, B. & Dwivedi, Y. (2018) Unfolded protein response and associated alterations in toll-like receptor expression and interaction in the hippocampus of restraint rats. Psychoneuroendocrinology, 89, 185-193. Torres-Peraza, J.F., Engel, T., Martín-Ibáñez, R., Sanz-Rodríguez, A., Fernandez-Fernandez, M.R., Esgleas, M., Canals, J.M., Henshall, D.C. & Lucas, J.J. (2013) Protective neuronal induction of ATF5 in endoplasmic reticulum stress induced by status epilepticus. Brain : a journal of neurology, 136, 1161-1176. van Dijk, R.M., Di Liberto, V., Brendel, M., Waldron, A.M., Möller, C., Gildehaus, F.J., von Ungern‐Sternberg, B., Lindner, M., Ziegler, S. & Hellweg, R.J.E. (2018) Imaging biomarkers of behavioral impairments: A pilot micro–positron emission tomographic study in a rat electrical post–status epilepticus model. 59, 2194-2205. Vaughan, D.N. & Jackson, G.D. (2014) The piriform cortex and human focal epilepsy. Front Neurol, 5, 259. Vavers, E., Svalbe, B., Lauberte, L., Stonans, I., Misane, I., Dambrova, M. & Zvejniece, L. (2017) The activity of selective sigma-1 receptor ligands in seizure models in vivo. Behav Brain Res, 328, 13-18. Vismer, M.S., Forcelli, P.A., Skopin, M.D., Gale, K. & Koubeissi, M.Z. (2015) The piriform, perirhinal, and entorhinal cortex in seizure generation. Front Neural Circuits, 9, 27. Walker, A., Russmann, V., Deeg, C.A., von Toerne, C., Kleinwort, K.J.H., Szober, C., Rettenbeck, M.L., von Ruden, E.L., Goc, J., Ongerth, T., Boes, K., Salvamoser, J.D., Vezzani, A., Hauck, S.M. & Potschka, H. (2016) Proteomic profiling of epileptogenesis in a rat model: Focus on inflammation. Brain, behavior, and immunity, 53, 138-158. Wang, M., Law, M.E., Castellano, R.K. & Law, B.K. (2018) The unfolded protein response as a target for anticancer therapeutics. Crit Rev Oncol Hematol, 127, 66-79. Wang, M., Wey, S., Zhang, Y., Ye, R. & Lee, A.S. (2009) Role of the unfolded protein response regulator GRP78/BiP in development, cancer, and neurological disorders. Antioxid Redox Signal, 11, 2307-2316. 32

Wang, S., Longo, F.M., Chen, J., Butman, M., Graham, S.H., Haglid, K.G. & Sharp, F.R. (1993) Induction of glucose regulated protein (grp78) and inducible heat shock protein (hsp70) mRNAs in rat brain after kainic acid seizures and focal ischemia. Neurochem Int, 23, 575-582. Welker Wally , J.J.I., Noe Adrianne, Comparative Mammalian Brain Collection - Cell Stain Brain Atlas of the Domestic Dog (Basenji) (Canis familiaris) #66-165. West, M.J., Slomianka, L. & Gundersen, H.J. (1991) Unbiased stereological estimation of the total number of neurons in thesubdivisions of the rat hippocampus using the optical fractionator. Anat Rec, 231, 482-497. Xiao, H. & Wu, R. (2017) Quantitative investigation of human cell surface N-glycoprotein dynamics. Chem Sci, 8, 268-277. Yamamoto, A., Murphy, N., Schindler, C.K., So, N.K., Stohr, S., Taki, W., Prehn, J.H. & Henshall, D.C. (2006) Endoplasmic reticulum stress and apoptosis signaling in human temporal lobe epilepsy. Journal of Neuropathology & Experimental Neurology, 65, 217-225. Zack, G.W., Rogers, W.E. & Latt, S.A. (1977) Automatic measurement of sister chromatid exchange frequency. J Histochem Cytochem, 25, 741-753. Zhang, G., Wang, X., Gillette, T.G., Deng, Y. & Wang, Z.V. (2019) Unfolded Protein Response as a Therapeutic Target in Cardiovascular Disease. Curr Top Med Chem. Zhu, X., Dong, J., Han, B., Huang, R., Zhang, A., Xia, Z., Chang, H., Chao, J. & Yao, H. (2017) Neuronal Nitric Oxide Synthase Contributes to PTZ Kindling EpilepsyInduced Hippocampal Endoplasmic Reticulum Stress and Oxidative Damage. Front Cell Neurosci, 11, 377. Zuo, D., Yu, X., Guo, C., Yi, H., Chen, X., Conrad, D.H., Guo, T.L., Chen, Z., Fisher, P.B. & Subjeck, J.R.J.T.F.J. (2012) Molecular chaperoning by glucose-regulated protein 170 in the extracellular milieu promotes macrophage-mediated pathogen sensing and innate immunity. 26, 1493-1505.

33

Legend to figures Fig 1: HSPA5 quantification in rat. Cell density of HSPA5 stereological cell count in the hippocampus (A: Hilus, B: CA1, C: CA3) and parahippocampal cortex (D: Entorhinal, E: Piriform-layer 1, F: Piriform-layer 2, G: Piriform-layer 3, H: Perirhinal). Data points correspond to cell density of the ipsi- and contralateral sides summed together. Representative images of the piriform cortex illustrating the increase of HSPA5 positive cells in SE (J, L) compared to CTR animals (I, K). Images are shown from animals with a history of a status epilepticus with sampling during the acute post-insult phase (I, J) and latency phase (K, L). All data are provided as mean (dashed line) ± SEM, * p < 0.05 (effect of electrical stimulation), # p < 0.05 (effect of time), two-way ANOVA. CTR: Electrode-implanted controls, SE: animals with history of status epilepticus, 2d: two days post-SE, 10d: ten days post-SE, 8w: eight weeks post-SE. Fig 2: HSPH4 quantification in rat. Cell density of HSPH4 stereological cell count in the hippocampus (A: Hilus, B: CA1, C: CA3) and parahippocampal cortex (D: Entorhinal, E: Piriform-layer 1, F: Piriform-layer 2, G: Piriform-layer 3, H: Perirhinal). Data points correspond to cell density of the ipsi- and contralateral sides summed together. Representative images of CA1 (I, M), piriform-layer 2 (J, N), entorhinal (K, O) and perirhinal (L, P) cortices showing the increase of HSPH4 positive cells in SE (M-P) compared to CTR animals (I-L). Images are shown from animals with a history of a status epilepticus with sampling during the latency phase. All data are given as mean (dashed line) ± SEM, * p < 0.05 (effect of electrical stimulation), # p < 0.05 (effect of time), two-way ANOVA. CTR: Electrode-implanted controls, SE: animals with history of status epilepticus 2d: two days post-SE, 10d: ten days post-SE, 8w: eight weeks post-SE. Fig 3: HSPA5 and HSPH4 quantification in dog. Immunopositive area respectively for HSPA5 and HSPH4 in hippocampal hilus (A, G), CA1 (B, K), CA3 (C, I), dentate gyrus (DG; D, J) regions and parahippocampal cortex (PHC; F, L). Data points correspond to averaged immunopositive area for HSPA5 and HSPH4 measured in each animal. Representative images of CA1 region staining for HSPA5 (E) and HSPH4 (H) in animals with structural epilepsy. All data are provided as mean (dashed line) ± SEM, * p < 0.05, one-way ANOVA. CTRexp: control dogs without neurological diseases used in parasitological studies. CTRpat: control dogs without neurological diseases treated in the clinic. Idiopathic: patients with idiopathic 34

epilepsy, subtype no identification of structural epilepsy and unknown cause. Structural: patients with epilepsy caused by identified cerebral pathology. Fig

4:

HSPA5

morphological

characterization

in

rat

brain.

Representative

microphotographs of double immunostaining of HSPA5 (A, D, G) with respectively NeuN (B), Iba1 (E) and GFAP (H). In the merged images (C, F, I), where also the Hoechst 33342 counterstain is visible in blue, the protein’s colocalization with a neuronal marker (C) but not with a microglia marker (F) is proofed. Colocalization of HSPA5 with GFAP is restricted to several cells in the layer 1 of the piriform cortex (I). Images are shown from animals with a history of a status epilepticus with sampling during the latency phase. Piriform cortex layer 3 (A-C), hilus of dentate gyrus (D-F) and piriform cortex layer 1 and 2 (G-I) are shown . Fig 5: HSPA5 expression in rat cortical astrocytes. A representative high magnification Zstack image demonstrating colocalization of HSPA5 and astroglial marker GFAP inside a cell in the piriform cortex layer 1. Several cells in the same region presented a similar expression pattern. The Z-stack image is shown from an animal with a history of a status epilepticus with sampling during the latency phase. Fig

6

HSPH4

morphological

characterization

in

rat

brain.

Representative

microphotographs of double immunostaining of HSPH4 (A, D, G) with NeuN (B), Iba1 (E) and GFAP (H) respectively. In the merged images (C, F, I), where also the Hoechst 33342 counterstain is visible in blue, the protein’s colocalization with a neuronal marker (C) but neither with a microglia (F), nor with an astrocyte marker (I) is proofed. Images are shown from animals with a history of a status epilepticus with sampling during the latency phase. Hilus of dentate gyrus (A-F) and Cornu Ammonis 3 (G-I) regions are presented.

Highlights 1. Expression of HSPA5 and HSPH4 is altered in rat hippocampus and piriform cortex. 2. The expression rates of HSPA5 and HSPH4 increase with time. 3. HSPA5 and HSPH4 are expressed in neurons; HSPA5 localizes in single astrocytes. 4. In canine patients with structural epilepsy, HSPH4 is up-regulated in CA1 region. 5. HSPA5 and HSPH4 may represent targets for anti-epileptogenic therapies. 35

36

37

38

39