1 Chronic levetiracetam decreases hippocampal ...

1 Chronic levetiracetam decreases hippocampal aromatase expression in normal but not temporal lobe epilepsy modeled rats and does not affect male reproductivity Mehtap Cincioğlu-Palabıyık a, Meral Buğdaycı b, Dilek Ertoy-Baydar b, Hande Karahan a, Yıldırım Sara c, Pelin Kelicen-Uğur.a* a

Affiliation address: Hacettepe University, Faculty of Pharmacy, Department of Pharmacology,

06100 Sıhhiye, Ankara, Turkey. Present address: Turkish Medicines and Medical Devices Agency (TITCK), Department of Regulatory Affairs, Division of Pharmacological Assessment, Ankara, Turkey. b

Hacettepe University, Faculty of Medicine, Department of Pathology, 06100 Sıhhiye, Ankara,

Turkey c

Hacettepe University, Faculty of Medicine, Department of Pharmacology, 06100 Sıhhiye,

Ankara, Turkey

*Correspondence to: Hacettepe University, Faculty of Pharmacy, Department of Pharmacology, 06100 Sıhhiye, Ankara, Turkey

ABSTRACT Reproductive disorders are more common in men with epilepsy taking anticonvulsant medications. Antiepileptic drugs and seizures in medial temporal lobe structures may cause gonadal dysfunction, including infertility, decreased libido and potency. This study investigated the possible effects of the highly efficient, new-generation antiepileptic drug (AED) levetiracetam (LEV) on central and gonadal aromatase expression and gonadal tissue functionality at 27 mg/kg/day (low dose) and increasing doses (54 and 108 mg/kg/day). A temporal lobe epilepsy (TLE) model was generated in male Wistar rats via intraperitoneal (i.p.) injection of the excitotoxic agent kainic acid (KA). Aromatase levels were 1.5 times higher in the brain cortex of the TLE groups after 4 weeks and hippocampus after 4 and 8 weeks compared to controls. Decreased basal aromatase levels were observed after 1 week of LEV treatment (27 mg/kg/day). Administration of low-dose LEV did not alter vas deferens contractions at 1, 4 or 8 weeks compared with controls. No histological changes were observed in vas deferens, epididymis or testis after 8 weeks of LEV administration in

2 increasing doses. Administration of 27 and 54 mg/kg/day LEV down-regulated testis aromatase expression at 8 weeks compared to controls. These results suggest that LEV decreases aromatase levels in the hippocampi at the beginning of the treatment as an antiepileptogenic effect and also decreases aromatase levels in the testis and increases the epileptic threshold via decreasing systemic estradiol levels. Key Words: Aromatase, kainic acid, levetiracetam, temporal lobe epilepsy, gonadal tissue. Chemical compounds studied in this article: (http://www.ncbi.nlm.nih.gov/pccompound) Levetiracetam (PubChem CID:441341); Carbachol (PubChem CID:5831); Phenylephrine (PubChem CID:6041); Kainic acid (PubChem CID:10255). Supported by H.Ü.B.A.B. (Project number 010 01 301 001).

3 1. Introduction Epilepsy and anticonvulsant medications may substantially alter endocrine homeostasis, including the male reproductive hormonal system. Gonadal dysfunction, including infertility, is commonly seen in men with epilepsy (Frye, 2010; Sivaraaman and Mintzer, 2011). The use of older generation medications that induce the cytochrome P450 system is associated with lower bioactive testosterone (Mintzer, 2010). Levels of circulating bioavailable testosterone are also affected by the aromatase enzyme, which converts testosterone to estrogen and may be affected by seizure medications (Reddy, 2004). Testosterone and gonadal functions are not affected by the newer generation (non-inducing) anticonvulsants (Xiaotian et al., 2013). However, the relationship of these drugs with central aromatase levels is not clear. Aromatase is a member of the p450 aromatase enzyme family, and it catalyzes estrogen biosynthesis from androgen precursors. Aromatase (P450 arom; CYP19) expression increases following brain trauma, experimental ischemic injury and neurodegeneration, including during the first stage of Alzheimer's disease. This increase suggests that aromatase plays an important role in the protection of brain nerve cells by increasing local estrogen levels (Garcia-Segura, 2008). However, the effect of estradiol in epilepsy is more complicated than neuroprotection. Estradiol is proconvulsant, and androstenediol and testosterone exhibit anticonvulsant properties (Woolley, 2000; Reddy, 2004). Estradiol benzoate treatment for amygdala-kindled seizures in male rats intensified seizures in a recent study (Saberi et al., 2001). Several of the so-called ‘newer generation’ antiepileptic drugs (AEDs) share one important property; these AEDs do not affect the CYP450 system. Levetiracetam (LEV) has recently gained wide acceptance for its unique antiepileptic mechanism, ideal pharmacokinetic properties, high efficacy and favorable safety profile (Xiaotian et al., 2013). LEV dose-dependently reduced the severity of kainic acid (KA)-induced seizures with significant protection at a dose of 54 mg/kg/day (i.p.) in rats (Loscher and Honack, 1993; Klitgaard et al., 1998). LEV binds to synaptic vesicle protein SV2A, which is prevalent in the central nervous system and most endocrine tissues, (Lynch et al., 2004; Vogl et al., 2012), but the mechanism by which LEV alters the reproductive endocrine balance is not fully understood. The effect of LEV on brain and gonadal aromatase expression is also not known. KA is an excitotoxic analogue of glutamate (Vincent and Mulle, 2009). Systemic or intracerebral administration of KA induces epileptic seizures in experimental animals (McLin and

4 Steward, 2006; Pereno and Beltramino, 2009) that closely resemble TLE in human (Ben-Ari, 1985; Zheng et al., 2011). Numerous studies using AEDs and selective ligands demonstrated the pharmacological properties of KA-induced seizures, but no systematic comparative study demonstrated the effects of LEV on aromatase expression in brain cortex, hippocampus and testis or histopathological changes in male gonadal tissues. This study investigated the effects of LEV on aromatase expression in brain cortex and hippocampus of male rats using the KA-induced TLE model. The effects of LEV on seladin-1 expression, which converts desmosterol to cholesterol, was also investigated. The central and gonadal expression of aromatase was investigated in rats using an extrapolated dose (27 mg/kg/day) of LEV from a clinically used human dose (1000 mg/75 kg adult) (Genton and Van Vleymen, 2000) and increasing doses (54 and 108 mg/kg/day). Contractile responses of vas deferens (27 mg/kg/day), aromatase expression in testis and the histological structure of gonadal tissues (27, 54 and 108 mg/kg/day) in male rats were also investigated at the different LEV doses.

2. Materials and methods

2.1. Materials Caspase-3 antibody (Cell Signaling Technology, Danvers, MA, USA), Hoescht stain (ThermoFisher Scientific, Waltham, MA, USA), HRP-GAM (Horseradish peroxidase-goat anti mouse) antibody, HRP-GAR (Horseradish peroxidase-goat anti rabbit) antibody, β-actin antibody, PVDF membrane, ECL (enhanced chemiluminescent) (Santa Cruz, Dallas, Texas, USA), photography film, developer-fixer (Kodak, Rochester, NY, USA), Lowry protein reactive solution A and B, MiniProtean TGX Precast gel, Laemmli 4X sample buffer, β-mercaptoethanol (BioRad, Hercules, CA, USA), Bovine serum albumin, Ethylenediaminetetraacetic acid, Magnesium chloride, Tris HCl, Tris base, Kainic acid, Sodium dodecyl sulphate, Tetramethylenediamine, Dithiothreitol, Sodium chloride, Glycerol, Phenylmethylsulfonyl fluoride, Acrylamidebisacrylamide, Carbachol, Phenylephrine (Sigma, St. Louis, MO, USA), Glycin (Applichem, Darmstadt, Germany), Protease inhibitor cocktail tablet (Roche Diagnostics, Mannheim, Germany), Aromatase antibody, (Acris, LA, USA), Tween-20, Entellan (Merck, Darmstadt, Germany), Skimmed milk powder (Pınar, İzmir, Turkey), Lidocaine (Sanovel, Istanbul, Turkey), Diazepam (Deva, Istanbul, Turkey), Ketamine, Xylazine (Pfizer, Cambridge, MA, USA), Dental

5 cement (Imicryl, Konya, Turkey), Cy-3-GAM antibody (Jackson Immunoresearch Lac., West Grove, PA), Levetiracetam [(S)-2-(2-Oxopyrrolidin-1-yl)butanamide] (Abdi Ibrahim, Istanbul, Turkey), Methanol, Bouin solution, Formaline, Lactated Ringer's solution.

2.2. Methods The Ethical Committee for Animal Experimentation of Hacettepe University approved this study (2012/26-13, 2014/30-2). Male adult Wistar rats weighing 200–250 g were obtained from the Hacettepe University Experimental Animal Unit, Turkey. Rats were housed in controlled conditions at a temperature of 22○C and a 12 h:12 h light–dark cycle (lights on at 7 a.m.). The animals were given free access to standard rat chow and water.

2.2.1. Electrode implantation and ECoG recording Electrodes were implanted under general anesthesia (90:10 mg/kg ketamine:xylazine, i.p.) one week before the initiation of experiments. Lidocaine was applied in rat ears prior to positioning the animal in a stereotaxic apparatus. A controlled heating pad and lamp above the rat were used to maintain body temperature within the normal range (37.0-37.5◦C) throughout the experiment and during surgery and the recovery period. Stainless-steel electrodes were permanently implanted stereotactically in the right frontoparietal cortex ∼3.5 mm right lateral and 2.0 and 4.0 mm rostral to the bregma on the cranium, and a reference electrode was implanted ∼1.0 mm left lateral and 2.5 mm rostral to the lambda on the cranium and 1.0 mm from the cortical surface, according to the brain atlas of Paxinos and Watson (Paxinos and Watson, 2004). Two screws were fixed on the parietal cortex, and one screw was fixed on the frontal cortex to support the electrodes. The electrode wires were attached to a connector, and the assembly was secured to the skull using dental acrylic cement. Paracetamol (0.1 mg/kg) and saline were injected intraperitoneally to prevent pain and dehydration, respectively. Animals were returned to their cages during recovery.

2.2.2. Kainic acid-induced seizures in rats Status epilepticus (SE) was induced according to a modified protocol based on the multiple, low-dose KA injection protocol of Hellier et al. (Hellier et al., 1998). This protocol is highly efficient, and it is associated with a low mortality rate. Rats were aged 6-8 weeks with an average

6 weight of 200-250 g during the SE induction stage. Animals received an initial dose of 5 mg/kg KA (i.p.), which was repeated hourly until SE was observed. All animals reached an SE that lasted for at least 4 hours after receiving an average dose of 7.5 mg/kg KA. Seizure activity was rated using the modified Racine’s Scale (Racine, 1972) of Hellier et al. (1998): Stage I, facial clonus; Stage II, nodding and wet dog shaking; Stage III, unilateral forelimb clonus with lordotic posture; Stage IV, lateral forelimb clonus with rearing; and Stage V, bilateral forelimb clonus with rearing, jumping, and falling. KA was injected for at least in 3 hours for the model to be fully established with SE with repetitive convulsive seizures. Recording continued during the KA treatment and for 3 hours. Diazepam (5 mg/kg; i.p.) was injected after the KA injection to prevent rat death due to SE. Lactated Ringer's solution (3 ml; subcutaneous) was injected in all KA-injected animals after this procedure to prevent dehydration. Rats were divided into 12 groups and treated as shown in Table 2.1. This experimental procedure lasted for 1 or 4 or 8 weeks as described in Fig. 2.1a. Six animal were used in each group, except the 8 week-Control (n=7) and 8 week-LEV (n=7) groups.

Table 2.1. Animal groups. Following days until Groups

1st experiment day

euthanasia

Group 1 (Control)

i.p. 0.9 % NaCl

-

Group 2 (Epileptic)

i.p. KA

i.p. 0.9 % NaCl

i.p. 0.9 % NaCl

i.p. LEV (27 mg/kg/day)

i.p. KA

i.p. LEV (27 mg/kg/day)

Group 3 (LEV) Group 4 (Treated)

Animal weights were recorded at the beginning and end of the study. ECoG was recorded for 3 hours in the chamber on the first day of the experiment. A microconnector was plugged into the ECoG recording system via cables with swivel junctions that allowed the animals to move freely within the chamber in Faraday cage. ECoG signals were amplified (×500), low-pass filtered and digitized at 1000 samples/sec. Analysis was performed off-line using software that was developed in the Pharmacology Laboratory of Hacettepe University, Faculty of Medicine. ECoGs after the first day of the experiment were recorded from all animals at 24 hours and 1, 4, 8 weeks for 20 minutes.

7

Figure 2.1. Schematic of experimental procedure

Rats were killed with carbon dioxide and decapitated at the designated time intervals. The cerebral cortices, the hippocampi and testes were isolated and stored at -80°C. Vas deferens were used in organ bath experiments. Some of the rats were treated chronically with LEV only (27, 54 and 108 mg/kg/day) without KA injection, as shown in Fig. 2b. Increasing doses of LEV (27, 54, 108 mg/kg/day) were administered over 8 weeks to three groups of rats. The control group received i.p. 0.9% NaCl. Therefore, four groups (n=3/group) were formed. These rats were killed via cardiac perfusion, and tissue isolation was performed immediately. Brains were maintained in paraformaldehyde, and right testis, vas deferens and epididymis were maintained in Bouin solution (Bouin et al., 2004).

2.2.3. Tissue lysis Brain cortex, hippocampus and testis for immunoblotting were homogenized on ice using specific lysis buffers in an ultrasonic homogenizer. Homogenates were centrifuged at 14,000×g at 4°C for 20 min. The protein concentration of each sample was assayed according to the method of Lowry (Lowry et al., 1951).

8 2.2.4. Western blotting Supernatants were added to Laemmli sample buffer/β-mercaptoethanol and boiled for 5 min. Equal amounts of protein per lane (100 µg) were loaded onto a 4-15% polyacrylamide gel and separated using SDS–PAGE under 300 V for 30 min in a Mini-Protean System (BioRad, USA). Proteins were transferred to PVDF membranes under 100 V for 90 min. Rat ovary extract was used as a positive control. Membranes were blocked with 5% non-fat milk diluted in TBS-T at room temperature for 2 h and incubated with aromatase and seladin-1 antibodies followed by secondary antibodies. Immunoreactive bands were visualized using an enhanced chemiluminescence system (ECL) followed by exposure of the membranes to autoradiographic films. The same blots were stripped and reprobed with a β-actin antibody, or the same samples were loaded onto a polyacrylamide gel, separated by electrophoresis, transferred to PVDF membranes and probed with a β-actin antibody as an internal control to ensure equal protein loading in all lanes. Band intensity was quantified using densitometry in Scion Image 4.0.3.2. (Scion Corp.), and data were normalized to β-actin as the protein control.

2.2.5. The effect of 27 mg/kg/day LEV on in vitro vas deferens contractile responses Animals were killed, and the vas deferens was isolated and cleaned from connective tissue around the muscle. Tissue segments 2-3 cm in length were mounted under a 1-g tension in a 10-ml organ bath that contained a Tyrode solution (Tyrode, 1910) bubbled with a 95% O2 - 5% CO2 mixture and maintained at 32°C. One ending of the vas deferens was attached to a force displacement, which was connected through a bridge amplifier to a PowerLab recording system (MP35 Transducer Data Acquisition System, Biopac, USA) that was coupled to a computer. The other ending was attached to a mounting hook. Tissues samples were stabilized for 30 min, and concentration–response curves were performed for phenylephrine and carbachol via the cumulative addition of these agonists to the organ bath. Changes in isometric tension were recorded.

2.2.6. Increasing doses (27, 54 and 108 mg/kg/day) of LEV on gonadal tissue histology Testis, vas deferens and epididymis were fixed in Bouin solution for 48 hours and subjected to overnight processing. Tissue sections were cut at 4-5-µm thickness from a microtome block. Sections underwent hematoxylin-eosin staining and were embedded in a synthetic resin and prepared for microscopic examination. Gonadal tissues were evaluated microscopically, and counts

9 and markings were made on the slides to simplify and improve the microscopic examination of tissues. Assessments were performed without skipping a homogeneous representation of the regions. The corpus diameters of epididymis and vas deferens were measured from rats that received LEV (27, 54, 108 mg/kg/day) for 8 weeks. Testis were examined and measurements were of the criteria mentioned below to determine whether tubules were functional or atrophic: 

The number of functional and atrophic tubules



The number of degenerated germ cell tubules



The number of obstructive tubules



Diameter of functional and atrophic tubules



Thickness of functional and atrophic tubules



Tubular interstitial space

2.3. Statistical analysis All tests were performed using Prism5 (GraphPad Prism Software Inc., USA), and all data were analyzed. The results are represented as the means  SEM. Significance versus control is designed as P < 0.05. Animal number per group is indicated with “n”. ECoG was recorded using Biopac Student Lab Analysis 3.7.7 (USA). A wave that exhibited an increase in the amplitude of cortex basal electrical activity of at least 2-fold was considered a spike. Spikes were counted empirically. One-way ANOVA followed Newman-Keuls multiple comparison tests were used to determine significant differences in seizure frequency (n=3-6). Blots were measured using ScionImage 4.0.3.2. (Scion Corp.), and data were analyzed using two-way ANOVA and Bonferroni's multiple comparison test. Contraction responses of vas deferens were analyzed using Biopac AcqKnowledge (USA), and Student’s t and Mann Whitney U tests were used for statistical analyses (n=6). Pathological measurements were performed using the Spot Advanced program, and results were analyzed using one-way ANOVA and Kruskal-Wallis tests (n=3).

3. Results

Rats were considered epileptic when they exhibited class III, IV and V seizures according to Racine’s Scale. Rats were placed in the recording chamber after the first injection to record KA-

10 induced electrographic epileptiform activity. All KA-injected animals underwent SE. These rats developed spontaneous generalized tonic and clonic seizures with a latent period of 3–5 days after SE.

3.1. LEV (27 mg/kg/day) decreased average spike number and amplitude in KA-injected rats ECoG recordings were examined in KA-administered groups. Average spike number per minute (Fig. 3.1) and average spike amplitude (Fig. 3.2) exhibited an increase at all time frames compared to controls. A significant decrease was observed in average spike number per minute in the KA and KA+LEV groups at all time frames compared to the first day of the experiment. There was also a significant decrease in average spike number per minute in KA+LEV (27 mg/kg/day) groups after 24 hours, 1 week, 4 weeks and 8 weeks compared to the matched KA groups (Fig. 3.1, n=6, P