- Email: [email protected]
Cerebral blood flow changes during pilocarpine-induced status epilepticus activity in the rat hippocampus
Experimental Neurology 225 (2010) 196–201
Contents lists available at ScienceDirect
Experimental Neurology j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / y e x n r
Cerebral blood flow changes during pilocarpine-induced status epilepticus activity in the rat hippocampus M. Choy a,b,c, J.A. Wells d, D.L. Thomas d, D.G. Gadian a, R.C. Scott a,c,e,f,1, M.F. Lythgoe a,b,⁎,1 a
Radiology and Physics Unit, UCL Institute of Child Health, University College London, London, UK Centre for Advanced Biomedical Imaging, UCL Department of Medicine and UCL Institute of Child Health, University College London, London, UK Neurosciences Unit, UCL Institute of Child Health, University College London, London, UK d Wellcome Trust Advanced MRI Group, Department of Medical Physics and Bioengineering, University College London, London, UK e Neurology Unit, Great Ormond Street Hospital, London, UK f National Centre for Young People with Epilepsy, Surrey, UK b c
a r t i c l e
i n f o
Article history: Received 9 March 2010 Revised 1 June 2010 Accepted 20 June 2010 Available online 25 June 2010 Keywords: Status epilepticus Hippocampus Cerebral blood flow MRI
a b s t r a c t Introduction: There is a known relationship between convulsive status epilepticus (SE) and hippocampal injury. Although the precise causes of this hippocampal vulnerability remains uncertain, potential mechanisms include excitotoxicity and ischaemia. It has been hypothesised that during the early phase of seizures, cerebral blood flow (CBF) increases in the cortex to meet energy demand, but it is unclear whether these compensatory mechanisms occur in the hippocampus. In this study we investigated CBF changes using perfusion MRI during SE in the pilocarpine rat. Methods: First, we determined whether SE could be induced under anaesthesia. Two anaesthetic protocols were investigated: isoflurane (n = 6) and fentanyl/medetomidine (n = 7). Intrahippocampal EEG electrodes were used to determine seizure activity and reflex behaviours were used to assess anaesthesia. Pilocarpine was administered to induce status epilepticus. For CBF measurements, MRI arterial spin labelling was performed continuously for up to 3 h. Either pilocarpine (375 mg/kg) (n = 7) for induction of SE or saline (n = 6) was administered. Diazepam (10 mg/kg) was administered i.p. 90 min after the onset of SE. Results and discussion: We demonstrated time-dependent significant (p b 0.05) differences between the CBF responses in the parietal cortex and the hippocampus during SE. This regional response indicates a preferential distribution of flow to certain regions of the brain and may contribute to the selective vulnerability observed in the hippocampus in humans. © 2010 Elsevier Inc. All rights reserved.
Introduction Convulsive status epilepticus (SE) is defined as a convulsive seizure that persists for 30 min or more (Rona et al., 2005) and has been associated with brain injury especially to the hippocampus (Scott et al., 2006; Choy et al., 2010). The mechanisms underlying this vulnerability of the hippocampus remain unclear. Seizure-induced excitotoxicity has been hypothesised to be the main mechanism that leads to injury, although there is some evidence that regional ischaemia may also be an important factor (Fabene et al., 2007; Siesjo and Wieloch, 1986). It has been suggested that there are acute augmentations (or increases) changes in cerebral blood flow (CBF) during a seizure (compensation phase), following which there is a reduction in CBF (decompensation phase) (for a review see Lothman, 1990). The
⁎ Corresponding author. Centre for Advanced Biomedical Imaging, Paul O'Gorman Building, University College London, 72 Huntley Street, London WC1E 6DD, UK. Fax: +44 20 7905 2358. E-mail address: [email protected] (M.F. Lythgoe). 1 These authors contributed equally to this manuscript. 0014-4886/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.expneurol.2010.06.015
compensation phase occurs within the first 30 min of a seizure and is characterised by recruitment of physiological mechanisms such as increased blood flow and alterations in pH and glucose metabolism (Lothman and Collins, 1981; Meldrum and Nilsson, 1976). Subsequently, the brain enters the decompensation phase during which several physiological processes, such as cerebral autoregulation, fail and neuronal death may begin to occur (Meldrum and Nilsson, 1976). Since autoregulation maintains local blood flow to the tissue, it has been hypothesised that a failure may result in local ischaemia that could exacerbate injury (Meldrum and Nilsson, 1976). It is possible that such a local ischaemia is part of the pathophysiological process that underlies the vulnerability of the hippocampus to status epilepticus induced injury. The ability to non-invasively image cerebral haemodynamics during seizures has been hampered by a lack of methods that facilitate the mapping of CBF with high temporal and spatial resolution, over a wide range of blood flows (Calamante et al., 1999). Arterial spin labelling (ASL) is an MRI technique for measuring CBF that uses magnetically labelled blood water as an endogenous tracer (Thomas et al., 2000). This technique does not have a limitation on the number of repeat measurements in a single study, as it does not
M. Choy et al. / Experimental Neurology 225 (2010) 196–201
require the injection of an exogenous tracer. This makes ASL well suited for the continuous monitoring of the timecourse of CBF over durations of minutes to hours, e.g. to monitor the heterogeneous evolution of tissue perfusion after experimental ischemia and reperfusion in animal models (Lythgoe et al., 2002). Therefore, it is ideal for evaluating the timecourse of cerebral blood flow changes associated with status epilepticus. The images generated with ASL can be converted into CBF maps (ml/100 g/min) and are particularly suited to measuring quantitative regional blood flow changes over the course of an epileptic seizure. Despite these benefits, ASL has yet to be used to investigate seizures directly although it has been used for investigating the pathology following seizures (Choy et al., 2010; Ndode-Ekane et al., 2010). Only a few studies have imaged the brain during seizure activity with MRI. Magnetic resonance diffusion-weighted imaging (DWI) has been shown to be sensitive we investigate whether to regional changes during seizures and most likely reflects microstructural changes in the tissue (van Eijsden et al., 2004; Engelhorn et al., 2007; Zhong et al., 1995). A recent study used BOLD fMRI to demonstrate that seizure activity in rats was associated with negative BOLD signals in hippocampus, indicating a regional hemodynamic change (Schridde et al., 2008). Perfusion MRI has been used in a previous study, which suggested that there is a relationship between the maximum decrease in perfusion and subsequent neuronal loss using a gadolinium-based contrast agent (Engelhorn et al., 2005). However, these techniques do not allow the characterisation of a timecourse and therefore regional blood flow changes as a function of time during the course of status epilepticus. In this study, we hypothesise that the brain haemodynamic responses to seizures change over time, with the greatest response occurring during the first 30 min of a seizure. We also hypothesise that the haemodynamic response of the hippocampus to seizures is impaired when compared to that of the cerebral cortex. Therefore, the aim of this study was to use continuous arterial spin labelling (CASL) in the pilocarpine model of status epilepticus in rats to characterise the regional blood flow changes from before the onset of SE to seizure termination.
197
rats (300–350 g) were randomly divided into two groups and anaesthetised with either isoflurane (1.5% in a 60/40 N2O:O2 mix delivered at 1 l/min) (n = 6) or fentanyl citrate (300 μg/kg, i.p) and medetomidine (300 μg/kg, i.p) (n = 7). Tungsten wire electrodes (50 μm diameter, Science Products GMBH, Germany) were implanted bilaterally at symmetrical points in the hippocampus for EEG recording (AP, −4 mm from bregma; ML, 2 mm, and DV, 3.2 mm from the neocortex). The hippocampal electrodes were referenced to a ground electrode attached to the tail of the animal. EEG was acquired continuously until the end of the experiment. Methylscopolamine (1 mg/kg, i.p., Sigma-Aldrich) was injected i. p. to reduce the peripheral cholinergic effects of pilocarpine, followed 15–20 min later by pilocarpine hydrochloride injection (375 mg/kg, i. p., Sigma-Aldrich). If seizure activity was detected on EEG, then 90 min after the onset of continuous seizure activity, diazepam (10 mg/kg, i.p., Phoenix Pharma Ltd, Gloucester, UK) was given to attenuate the seizure. Pilocarpine hydrochloride and methylscopolamine were freshly prepared prior to administration in 0.9% saline. Reflex behaviour was used to assess the level of anaesthesia. Two reflex behaviours were used to provide an estimation of the depth of anaesthesia required for surgical procedures: paw pinch reflex and corneal reflex (Alves et al., 2009; Green, 1982). The depth of anaesthesia was tested every 10 min using these reflex behaviours to determine whether adequate anaesthesia was maintained throughout the experiment. If the animal retained either reflex during any period of the protocol then, for fentanyl, an additional dose of 100 μg/kg of was given, or for isoflurane, the inspired dose was increased. The presence of status epilepticus was investigated using EEG which was recorded continuously using a Bioamp differential amplifier interfaced with a Powerlab data acquisition system (AD Instruments, Australia). Signals were filtered between 1 and 120 Hz, a 50-Hz notch filter was used to reduce noise, with a sampling rate of 1024 samples per second. Analysis of electrophysiological data was carried out off-line on a Pentium computer, using Chart software (AD Instruments, Australia).
Methods All animal care and procedures were carried out in accordance with the UK Animals (Scientific Procedures) 1986 Act. As MR imaging of small animals typically requires the use of anaesthesia, an initial study was performed to determine whether pilocarpine can be used to induce status epilepticus under anaesthesia. Following these initial experiments, we investigated the regional nature of CBF changes during pilocarpine-induced SE using ASL. Study 1. Pilocarpine-induced SE under anaesthesia Anaesthesia has been shown to be effective for abolishing seizure activity, and used clinically in seizures refractory to conventional anticonvulsant therapy (Shorvon, 1994). However, it is commonly required in MR imaging of small animals as anaesthesia reduces stress and limits motion, which can lead to artefacts in the acquired images (Lukasik and Gillies, 2003). It was therefore necessary to determine that seizures could be induced with pilocarpine under anaesthesia. Isoflurane is a commonly used gaseous anaesthetic for MR imaging, due to, in part, its rapid clearance and ease of management for the animal in the MR system (Lukasik and Gillies, 2003), although it is known to be an effective anti-convulsant (Shorvon, 1994). Fentanyl, by itself, has been shown to be an effective method for sedation and subsequent induction of seizures (Engelhorn et al., 2005), but does not meet the requirement for anaesthesia (Flecknell, 2009). However, fentanyl in combination with medetomidine, can provide effective anaesthesia (Flecknell, 2009). Therefore we investigated the ability of pilocarpine to induce SE under these two anaesthetic regimens. Thirteen adult male Sprague–Dawley
Study 2. CBF measurements with MRI 19 adult male Sprague–Dawley rats (300–400 g, Charles Rivers) were divided randomly into two groups: pilocarpine (n = 13 of which 6 were discarded because of movement artefacts in the images) or saline (n = 6). Once the animals had been prepared for imaging (see below), scopolamine (1 mg/kg) was administered. T1-weighted images were acquired at the beginning of the experiment for the purpose of CBF quantification. Thereafter ASL was acquired continuously for the duration of the experiment. The experimental design is outlined in Fig. 1. The first 30 min of MRI were acquired for baseline measurements after which the animals were administered either pilocarpine (375 mg/kg) or saline. Bench experiments (see study 1) indicated that SE began approximately 30 min following pilocarpine administration; therefore in order to obtain data from 2 h of continuous seizure activity the animals were imaged for 150 min before diazepam (10 mg/kg) was given. After diazepam the rats were imaged for up to a further 30 min. The animals were visually assessed for seizure activity following each set of perfusion acquisitions. Animals were anaesthetised with intraperitoneal injections of fentanyl citrate (300 μg/kg) and medetomidine (300 μg/kg). Additional doses of fentanyl (100 μg/kg) were given if the paw pinch reflex was present in order to maintain anaesthesia. O2 was delivered continuously via a nose cone at a rate of 1 l/min. Rats were placed on a speciallydesigned animal holder to minimize motion artifacts. Physiological monitoring included electrocardiography (ECG) recordings and rectal temperature recordings. Temperature was maintained at 37 °C± 2.
198
M. Choy et al. / Experimental Neurology 225 (2010) 196–201
Fig. 1. Experimental design for MRI-ASL measurements during SE. The key events that occur over the course of MR imaging are as follows: 30 min of baseline acquisition before pilocarpine administration at time 0, SE onset occurs approximately 30 min after administration and diazepam was given 120 min after pilocarpine.
Images were acquired using a volume transmitter coil and a separate decoupled surface receiver coil. A central coronal imaging slice was selected that included the hippocampus (−4.8 mm from bregma). The imaging slice was determined by visual inspection of anatomical landmarks. All sequences were acquired from a 2-mmthick coronal slice. A low average perfusion acquisition was used in this study to reduce the effects of motion. For non-invasive CBF measurement, we used the continuous arterial spin labelling (CASL) method (Alsop and Detre, 1996), based on spin-echo echo-planar imaging (EPI) with interleaved adiabatic fast-passage inversion and control measurements. The centre of the inversion plane was situated 2 mm behind the cerebellum in all animals (approximately 10 kHz frequency offset); for the control image the same offset value was used, but with the sign reversed. The CASL labelling pulse duration was 3 s, and a post-labelling delay time of 500 ms was used to minimise transit time effects and intravascular artefacts (Thomas et al., 2006). The echo time was 36 ms, the interexperiment delay was 1000 ms, and 22 averages were acquired. A non-slice-selective inversion recovery (IR) EPI sequence was used to obtain the tissue T1 and the spin density (M0) parameters necessary for subsequent CBF quantification. IR-EPI images were acquired with: TR = 2000 ms, TE = 36 ms, an inversion time (TI) array of 284 ms, 434 ms, 634 ms, 834 ms, 1234 ms, 1734 ms, 2734 ms, and 3734 ms, and 22 averages were used. The IR data were fitted to obtain T1, α0 and M0 assuming a mono-exponential recovery function. Subsequently, these values were used to calculate CBF from the difference between the labelled and control CASL images (Alsop and Detre, 1996). The following assumptions were used to determine CBF: T1 of arterial
blood = 1.5 s, blood:brain partition coefficient for water (λ) = 0.9, efficiency of the spin labelling pulse = 0.71, tissue transit time = 250 ms, arterial transit time = 200 ms. Durations for each acquisition were: T1 = 11 min and CASL = 4 min. Each animal was scanned for a total time of 3 h. T1 maps were reconstructed using in-house software written in Matlab version 6.5 (MathWorks, Massachusetts, USA). As motion artefacts may lead to CBF quantification errors, perfusion-weighted images were visually inspected, and images with visible motion artefacts were excluded from the study. However, removal of these images decreased the signal to noise of these images, and therefore limited the precision of CBF measurements. We therefore averaged the images across 5 distinct periods to achieve reasonable precision for CBF, which were determined by the EEG experiments described in study 1 and from previous studies reviewed by Lothman (Lothman, 1990). These were: 1. baseline. 2. Within the first 30 min following pilocarpine administration. This is the period before SE occurs. 3. 30–60 min after pilocarpine. From our previous experiments, this period coincides with the first 30 min of SE and has been associated with the compensation phase of SE. 4. 60–150 min after pilocarpine. This period is associated with the latter stages of SE and this is considered to be the decompensation phase. 5. After approximately 150 min, diazepam was administered to the animal, and this period should reflect the gradual cessation of epileptic activity.
Fig. 2. Representative intrahippocampal EEG recording in a fentanyl/medetomidine anaesthetised rat. (a) 3 h EEG acquisition for the duration of experiment; (b) baseline EEG; (c) EEG desynchronisation post-pilocarpine (375 mg/kg); (d) increase in EEG indicating status epilepticus; (e) decrease in EEG following diazepam (10 mg/kg). Black arrow denotes pilocarpine administration, open arrow indicates diazepam administration. Data presented as mean ± SEM.
M. Choy et al. / Experimental Neurology 225 (2010) 196–201
199
Table 1 Mean regional CBF measurements before, during, and after pilocarpine. Regions include the parietal cortex, hippocampus and the thalamus. Data presented as mean ± SEM. CBF (ml/100 g/min)
Parietal cortex
Hippocampus
Thalamus
Baseline 0–30 min 30–60 min 60–120 min 120–150 min
122 ± 11 234 ± 32 121 ± 24 132 ± 32 165 ± 24
96 ± 14 118 ± 21 65 ± 19 105 ± 15 90 ± 27
109 ± 15 155 ± 24 108 ± 12 153 ± 30 150 ± 28
Three regions of interest (parietal cortex, hippocampus and thalamus) were delineated on all processed images (Fig. 3) and analysed with inhouse Matlab software and ImageJ version 1.38x (National Institutes of Health, USA). Statistical analysis For CBF measurements, a 3-way repeated measures ANOVA was conducted with contrasts to the baseline scans and the hippocampus. The main effects were for group (pilocarpine, saline), time (baseline, pilocarpine, SE, SE late, diazepam) and anatomy (parietal cortex, hippocampus, thalamus). All data presented as mean ± standard error of the mean (sem).
Results
Fig. 4. Ratios of parietal cortex CBF to hippocampal CBF over the course of SE. The periods are divided up as follows: 0–30 min is the period following pilocarpine injection; 30– 120 min during SE and; 120–150 min is following the injection of diazepam. Dashed line with open diamonds is the saline treated rats (n = 6), and full line with filled squares is the pilocarpine treated rats (n= 7). Data presented as mean± SEM, and significant statistical differences are indicated by * (p b 0.05).
previous reports in this model without anaesthesia (83%) (Mello et al., 1993) (Fig. 2d). Of the animals that did progress onto SE, the mean SE onset time following pilocarpine administration was 22.8 ± 1.9 min. In these animals, seizure activity persisted until the administration of diazepam, at which point the amplitude of the EEG gradually reduced (Fig. 2e).
Induction of SE under anaesthesia CBF measurements Throughout the duration of the experiments no reflex behaviours were observed using either anaesthetic regime. Following isofluraneanaesthesia, animals did not show any sign of seizure behaviour following pilocarpine. However, the fentanyl/medetomidine anaesthetised animals displayed facial automatisms and tonic extensions of the forelimbs and hindlimbs, in which the entire body of the rat became outstretched. These tonic extensions lasted for a few seconds, occurred intermittently, and only during the period of increased EEG activity. Fentanyl/medetomidine animals (6/7) displayed all of these behavioural responses. These effects are the main reason for the movement artefact that we removed from our data. Both groups of anaesthetised animals displayed a reduction in EEG amplitude following pilocarpine administration. The reduced amplitude persisted for the duration of the experiments in the isofluraneanesthetised animals. In the fentanyl/medetomidine-anaesthetised animals (Fig. 2), the reduced amplitude continued until the onset of SE. None of the animals under isoflurane anaesthesia progressed to SE (n = 6). In contrast, 6 out of 7 (85%) animals under fentanyl/ medetomidine anaesthesia progressed to SE, which is similar to
In total, 7 animals were included for analysis from the pilocarpine group (Table 1). In the control group, the estimated CBF remained relatively stable following saline injection (n = 6). Following pilocarpine injection, we observed an increase in parietal cortex blood flow relative to hippocampal blood flow at 0–30 min (Fig. 3.) (F = 14.9, p b 0.01) and at 30–60 min (F = 7.4, p = 0.02), but not after that time. The cortical blood flow was 2.84 ± 0.5 times more than hippocampal blood flow 30–60 min (Fig. 4). However, in the saline control group the cortical blood flow was only 1.46 ± 0.2 times more than hippocampal blood flow 30–60 min (F = 6.87, p = 0.03) Furthermore the hippocampus was the only region in which the mean blood flow decreased below the mean baseline level (baseline = 96 ± 14 ml/min/100 g; 30–60 min = 65 ± 19 ml/min/100 g). No significant differences were found between parietal cortex and thalamus or hippocampus and thalamus. These data indicate hippocampus and the parietal cortex respond to pilocarpine in a different manner and that the hippocampus is hypoperfused compared to the parietal cortex (Table 2).
Fig. 3. Representative mean perfusion-weighted images during SE from (a) a saline-injected and (c) a pilocarpine-injected animal. Panel b shows a T1 map use to calculate quantitative blood flow and the ROIs are outlined. The regions are: (1) parietal cortex; (2) hippocampus; (3) thalamus. Note the limited perfusion change in the hippocampus compared to the cortex. Data presented as mean ± SEM.
200
M. Choy et al. / Experimental Neurology 225 (2010) 196–201
Table 2 Regional and time-dependent CBF differences before, during, and after pilocarpine. Note that significant differences were found between the hippocampus and the parietal cortex at 0–30 and 30–60 min after pilocarpine injection when compared to baseline.
Thalamus vs. Hippocampus
Parietal cortex vs. Hippocampus
Time period contrast with baseline (min)
F
p
0–30 30–60 60–120 120–150 0–30 30–60 60–120 120–150
0.39 0.50 0.07 0.07 14.92 7.35 3.05 0.72
0.54 0.49 0.80 0.80 b 0.01 0.02 0.11 0.42
Discussion There were two main findings from this ASL-MRI study. Firstly we show that pilocarpine can be used to induce SE under fentanyl/ medetomidine anaesthesia, and that seizure activity occurs in the hippocampus. Secondly, during the course of pilocarpine-induced SE, cortical perfusion is high relative to hippocampal perfusion. An important component of the current study is the development of a model in which seizure activity could be induced without the behavioural manifestations of SE, such as the convulsive movement, confounding the experiments. We were unable to achieve this using isoflurane as an anaesthetic regime as no seizure activity was observed. As fentanyl with medetomidine has previously been shown to elicit anaesthesia suitable for surgical procedures in rodents (Meert and De, 1994; Wolfensohn and Lloyd, 1998) we investigated this anaesthetic protocol and confirmed that pilocarpine could be used to induce SE under this regime and therefore we were able to investigate cerebral blood flow associated with SE. Whether fentanyl/medetomidine anaesthesia has an effect on the elicited seizure remains unclear. The classes of drugs that fentanyl and medetomodine fall under, opiates and α2-adrenergic agonists, respectively, have been shown to affect seizure activity. Opiates have been found to potentiate seizures (Cherng and Wong, 2005), and α2adrenergic agonists have been shown to reduce seizure threshold (Mirski et al., 1994) as well as suppress seizure activity (Halonen et al., 1995). However, the interaction between these two classes of drug on seizures has not been well documented. Comparing our data with nonanaesthetised pilocarpine-induced seizures, we observed similar seizure onset times as well as a similar number of animals that progress on to SE (Cavalheiro et al., 2006). Also we demonstrate continued hippocampal seizure activity for at least 90 min, and that the seizure activity was attenuated following diazepam administration, indicating similarities between the pilocarpine models. Despite these similarities, both anaesthesia and pilocarpine affect central neurotransmitter systems and at some level induced seizures may be affected by anaesthesia, which will require further investigation. However, for the purposes of this study of investigating the time-dependent and regional CBF changes, the primary goal was eliciting prolonged hippocampal seizure activity, which was achievable in anaesthetised animals. Previous studies that have investigated the regional CBF profile during SE have typically used autoradiography techniques, which can provide regional CBF measurements that are limited to a single time point, or electrochemical techniques that provide continuous CBF measurements at a single small anatomical point in the brain (Meldrum and Nilsson, 1976; Shih and Scremin, 1992; Tanaka et al., 1990; Kreisman et al., 1991; Andre et al., 2002; Nersesyan et al., 2004). Therefore data providing whole regional CBF changes over time during SE are still required. We used ASL-MRI to investigate the CBF time course before, during and after pilocarpine-induced SE across a number of brain structures. Repeated ASL measurements demonstrated an initial CBF increase following pilocarpine administration in the three regions investigated.
The data provide good evidence for a reduced haemodynamic response in the hippocampus when compared to the cortex. A similar pattern was also observed in the thalamus relative to the hippocampus although the degree of perfusion change in the cortex tended to be greater than the thalamus. Absolute values of CBF indicated that the flow was not constant during the seizure process, which may represent loss of cerebral autoregulation after 30 min of seizure activity (Lothman, 1990). Our hippocampal EEG measurements indicate that seizure activity occurs continuously in the hippocampus under our anaesthetic regimen and persists until diazepam administration; therefore seizure activity alone is unlikely to account for these observed CBF variations. One drawback in our study was the time resolution, which limits a full assessment of the hemodynamic changes, and a further characterisation is necessary to elucidate the precise nature of prolonged seizure activity and the absolute CBF time-dependent changes. Our data demonstrate a region-specific response during seizure activity. The changes in parietal cortical CBF are consistent with previous studies during the periods of seizure activity. Accompanying increases in glucose utilisation and CMRO2 have also been reported (Shih and Scremin, 1992; Andre et al., 2002; Pinard et al., 1984; Tanaka et al., 1990; Kreisman et al., 1991). These data suggest that increases in CBF act to compensate for the increase in nutrient demand that the activated neurons require in the early phase of seizures. We observed that the hippocampus was hypoperfused compared to the cortex, which is endorsed by previous studies investigating bicuculline-induced seizures, where the authors have show a negative BOLD response in the hippocampus which they attributed to a mismatch between CMRO2 and CBF (Schridde et al., 2008; DeSalvo et al., 2010). An autoradiography study using a kainic acid seizure model also reported a limited hippocampal CBF response that was insufficient for the increase in metabolic demand (Tanaka et al., 1990), which has also been observed in the soman model of SE (Shih and Scremin, 1992). Taken together, these data suggest that the blood flow to the hippocampus may not meet local increases in energy demand during seizure activity, leading to a regional ischaemia. This may reflect the later period of SE for which there is a hypothesised failure of cerebral autoregulation, and may explain the reduction in blood flow we observed. Nevertheless contradictory reports exist. Substantial increases in hippocampal flow relative to the cortex have also been reported in the kainic acid model using helium clearance, an effect not observed in the bicuculline model (Pinard et al., 1987). This may suggest that the method of seizure induction may have an effect on the haemodynamic response to prolonged seizures. Whether the method of seizure induction with anaesthesia has or has not influenced the haemodynamic response is unclear, however, the non-anaesthetised soman model used by Shih and Scremin acts, like pilocarpine, to enhance cholinergic activity also reported a blunted CBF response, suggesting that this may be a property of the induced seizures (Shih and Scremin, 1992). These data therefore suggest that a relative hypoperfusion may contribute to injury during SE, which may have implications for treatment of SE in humans. In conclusion, this study indicates that regional CBF changes occur in vivo during pilocarpine-induced SE under fentanyl/medetomidine anaesthesia. When compared to cortical perfusion, the hippocampus CBF was consistently lower and these data support a hypothesis that this reduced blood flow may play a role in the selective vulnerability of the hippocampus to SE, which may, at least in part, underlie the seizureinduced hippocampal damage observed in animals and humans. Acknowledgments This work was supported by Epilepsy Research UK, Wellcome Trust, the Biotechnology and Biological Sciences Research Council and the British Heart Foundation. The authors thank Dr. Martin King for his advice on statistics, and Prof. Matthew Walker for his guidance on the pilocarpine model.
M. Choy et al. / Experimental Neurology 225 (2010) 196–201
References Alsop, D.C., Detre, J.A., 1996. Reduced transit-time sensitivity in noninvasive magnetic resonance imaging of human cerebral blood flow. J. Cereb. Blood Flow Metab. 16, 1236–1249. Alves, H.C., Valentim, A.M., Olsson, I.A., Antunes, L.M., 2009. Intraperitoneal anaesthesia with propofol, medetomidine and fentanyl in mice. Lab Anim. 43, 27–33. Andre, V., Henry, D., Nehlig, A., 2002. Dynamic variations of local cerebral blood flow in maximal electroshock seizures in the rat. Epilepsia 43, 1120–1128. Calamante, F., Thomas, D.L., Pell, G.S., Wiersma, J., Turner, R., 1999. Measuring cerebral blood flow using magnetic resonance imaging techniques. J. Cereb. Blood Flow Metab. 19, 701–735. Cavalheiro, E.A., Naffah-Mazzacoratti, M.G., Mello, L.E., Leite, J.P., 2006. Chapter 35, Models of Seizures and Epilepsy. Elsevier Inc. Cherng, C.H., Wong, C.S., 2005. Effect of fentanyl on lidocaine-induced convulsions in mice. Pharmacology 75, 1–4. Choy, M., Cheung, K.K., Thomas, D.L., Gadian, D.G., Lythgoe, M.F., Scott, R.C., 2010. Quantitative MRI predicts status epilepticus-induced hippocampal injury in the rat lithium-pilocarpine model. Epilepsy Res. 88 (2–3), 221–230. Desalvo, M.N., Schridde, U., Mishra, A.M., Motelow, J.E., Purcaro, M.J., Danielson, N., Bai, X., Hyder, F., Blumenfeld, H., 2010. Focal BOLD fMRI changes in bicuculline-induced tonic–clonic seizures in the rat. Neuroimage 50 (3), 902–909. Engelhorn, T., Doerfler, A., Weise, J., Baehr, M., Forsting, M., Hufnagel, A., 2005. Cerebral perfusion alterations during the acute phase of experimental generalized status epilepticus: prediction of survival by using perfusion-weighted MR imaging and histopathology. AJNR Am. J. Neuroradiol. 26, 1563–1570. Engelhorn, T., Weise, J., Hammen, T., Bluemcke, I., Hufnagel, A., Doerfler, A., 2007. Early diffusion-weighted MRI predicts regional neuronal damage in generalized status epilepticus in rats treated with diazepam. Neurosci. Lett. 417, 275–280. 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. Flecknell, P., 2009. Laboratory animal anaesthesia, Third Edition. Elsevier Inc. Green, C.J., 1982. Animal anaesthesia. Laboratory Animals Ltd. Halonen, T., Kotti, T., Tuunanen, J., Toppinen, A., Miettinen, R., Riekkinen, P.J., 1995. Alpha 2-adrenoceptor agonist, dexmedetomidine, protects against kainic acidinduced convulsions and neuronal damage. Brain Res. 693, 217–224. Kreisman, N.R., Magee, J.C., Brizzee, B.L., 1991. Relative hypoperfusion in rat cerebral cortex during recurrent seizures. J. Cereb. Blood Flow Metab. 11, 77–87. Lothman, E., 1990. The biochemical basis and pathophysiology of status epilepticus. Neurology 40, 13–23. Lothman, E.W., Collins, R.C., 1981. Kainic acid induced limbic seizures: metabolic, behavioral, electroencephalographic and neuropathological correlates. Brain Res. 218, 299–318. Lukasik, V.M., Gillies, R.J., 2003. Animal anaesthesia for in vivo magnetic resonance. NMR Biomed. 16, 459–467. Lythgoe, M.F., Thomas, D.L., Calamante, F., 2002. MRI measurement of cerebral perfusion and the application to experimental neuroscience. In: van-Bruggen, N., Roberts, T.P. (Eds.), Biomedical Imaging in Experimental Neuroscience. CRC Press, Boca Raton, Florida. Meldrum, B.S., Nilsson, B., 1976. Cerebral blood flow and metabolic rate early and late in prolonged epileptic seizures induced in rats by bicuculline. Brain 99, 523–542.
201
Meert, T.F., De, K.M., 1994. Potentiation of the analgesic properties of fentanyllike opioids with alpha 2-adrenoceptor agonists in rats. Anesthesiology 81, 677–688. Mello, L.E., Cavalheiro, E.A., Tan, A.M., Kupfer, W.R., Pretorius, J.K., Babb, T.L., Finch, D.M., 1993. Circuit mechanisms of seizures in the pilocarpine model of chronic epilepsy: cell loss and mossy fiber sprouting. Epilepsia 34 (6), 985–995. Mirski, M.A., Rossell, L.A., McPherson, R.W., Traystman, R.J., 1994. Dexmedetomidine decreases seizure threshold in a rat model of experimental generalised epilepsy. Anesthesiology 81, 1422–1428. Ndode-Ekane, X.E., Hayward, N., Gröhn, O., Pitkänen, A., 2010. Vascular changes in epilepsy: functional consequences and association with network plasticity in pilocarpine-induced experimental epilepsy. Neuroscience 166 (1), 312–332. Nersesyan, H., Herman, P., Erdogan, E., Hyder, F., Blumenfeld, H., 2004. Relative changes in cerebral blood flow and neuronal activity in local microdomains during generalized seizures. J. Cereb. Blood Flow Metab. 24, 1057–1068. Pinard, E., Tremblay, E., Ben-Ari, Y., Seylaz, J., 1984. Blood flow compensates oxygen demand in the vulnerable CA3 region of the hippocampus during kainate-induced seizures. Neuroscience 13, 1039–1049. Pinard, E., Rigaud, A.S., Riche, D., Naquet, R., Seylaz, J., 1987. Continuous determination of the cerebrovascular changes induced by bicuculline and kainic acid in unanaesthetized spontaneously breathing rats. Neuroscience 23, 943–952. Rona, S., Rosenow, F., Arnold, S., Carreno, M., Diehl, B., Ebner, A., Fritsch, B., Hamer, H.M., Holthausen, H., Knake, S., Kruse, B., Noachtar, S., Pieper, T., Tuxhorn, I., Luders, H.O., 2005. A semiological classification of status epilepticus. Epileptic Disord. 7, 5–12. Schridde, U., Khubchandani, M., Motelow, J.E., Sanganahalli, B.G., Hyder, F., Blumenfeld, H., 2008. Negative BOLD with large increases in neuronal activity. Cereb. Cortex 18, 1814–1827. Scott, R.C., King, M.D., Gadian, D.G., Neville, B.G., Connelly, A., 2006. Prolonged febrile seizures are associated with hippocampal vasogenic edema and developmental changes. Epilepsia 47, 1493–1498. Shih, T.M., Scremin, O.U., 1992. Cerebral blood flow and metabolism in soman-induced convulsions. Brain Res. Bull. 28, 735–742. Shorvon, S., 1994. Status Epilepticus. Cambridge University Press. Siesjo, B.K., Wieloch, T., 1986. Epileptic brain damage: pathophysiology and neurochemical pathology. Adv. Neurol. 44, 813–847 813-847. Tanaka, S., Sako, K., Tanaka, T., Nishihara, I., Yonemasu, Y., 1990. Uncoupling of local blood flow and metabolism in the hippocampal CA3 in kainic acid-induced limbic seizure status. Neuroscience 36, 339–348. Thomas, D.L., Lythgoe, M.F., Calamante, F., Pell, G.S., Ordidge, R.J., 2000. The measurement of diffusion and perfusion in biological systems using magnetic resonance imaging. Phys. Med. Biol. 45, R97–R138. Thomas, D.L., Lythgoe, M.F., van der Weerd, L., Ordidge, R.J., Gadian, D.G., 2006. Regional variation in cerebral blood flow and arterial transit time in the normal and hypoperfused rat brain measured using continuous arterial spin labeling MRI. J. Cereb. Blood Flow Metab. 26, 274–282. van Eijsden, P., Notenboom, R.G., Wu, O., de Graan, P.N., van Nieuwenhuizen, O., Nicolay, K., Braun, K.P., 2004. In vivo (1)H magnetic resonance spectroscopy, T(2)weighted and diffusion-weighted MRI during lithium-pilocarpine-induced status epilepticus in the rat. Brain Res. 1030, 11–18. Wolfensohn, S., Lloyd, M., 1998. Handbook of laboratory animal management and welfare, Second Ed. Blackwell Publishing. Zhong, J., Petroff, O.A., Prichard, J.W., Gore, J.C., 1995. Barbiturate-reversible reduction of water diffusion coefficient in flurothyl-induced status epilepticus in rats. Magn. Reson. Med. 33, 253–256.
Contents lists available at ScienceDirect
Experimental Neurology j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / y e x n r
Cerebral blood flow changes during pilocarpine-induced status epilepticus activity in the rat hippocampus M. Choy a,b,c, J.A. Wells d, D.L. Thomas d, D.G. Gadian a, R.C. Scott a,c,e,f,1, M.F. Lythgoe a,b,⁎,1 a
Radiology and Physics Unit, UCL Institute of Child Health, University College London, London, UK Centre for Advanced Biomedical Imaging, UCL Department of Medicine and UCL Institute of Child Health, University College London, London, UK Neurosciences Unit, UCL Institute of Child Health, University College London, London, UK d Wellcome Trust Advanced MRI Group, Department of Medical Physics and Bioengineering, University College London, London, UK e Neurology Unit, Great Ormond Street Hospital, London, UK f National Centre for Young People with Epilepsy, Surrey, UK b c
a r t i c l e
i n f o
Article history: Received 9 March 2010 Revised 1 June 2010 Accepted 20 June 2010 Available online 25 June 2010 Keywords: Status epilepticus Hippocampus Cerebral blood flow MRI
a b s t r a c t Introduction: There is a known relationship between convulsive status epilepticus (SE) and hippocampal injury. Although the precise causes of this hippocampal vulnerability remains uncertain, potential mechanisms include excitotoxicity and ischaemia. It has been hypothesised that during the early phase of seizures, cerebral blood flow (CBF) increases in the cortex to meet energy demand, but it is unclear whether these compensatory mechanisms occur in the hippocampus. In this study we investigated CBF changes using perfusion MRI during SE in the pilocarpine rat. Methods: First, we determined whether SE could be induced under anaesthesia. Two anaesthetic protocols were investigated: isoflurane (n = 6) and fentanyl/medetomidine (n = 7). Intrahippocampal EEG electrodes were used to determine seizure activity and reflex behaviours were used to assess anaesthesia. Pilocarpine was administered to induce status epilepticus. For CBF measurements, MRI arterial spin labelling was performed continuously for up to 3 h. Either pilocarpine (375 mg/kg) (n = 7) for induction of SE or saline (n = 6) was administered. Diazepam (10 mg/kg) was administered i.p. 90 min after the onset of SE. Results and discussion: We demonstrated time-dependent significant (p b 0.05) differences between the CBF responses in the parietal cortex and the hippocampus during SE. This regional response indicates a preferential distribution of flow to certain regions of the brain and may contribute to the selective vulnerability observed in the hippocampus in humans. © 2010 Elsevier Inc. All rights reserved.
Introduction Convulsive status epilepticus (SE) is defined as a convulsive seizure that persists for 30 min or more (Rona et al., 2005) and has been associated with brain injury especially to the hippocampus (Scott et al., 2006; Choy et al., 2010). The mechanisms underlying this vulnerability of the hippocampus remain unclear. Seizure-induced excitotoxicity has been hypothesised to be the main mechanism that leads to injury, although there is some evidence that regional ischaemia may also be an important factor (Fabene et al., 2007; Siesjo and Wieloch, 1986). It has been suggested that there are acute augmentations (or increases) changes in cerebral blood flow (CBF) during a seizure (compensation phase), following which there is a reduction in CBF (decompensation phase) (for a review see Lothman, 1990). The
⁎ Corresponding author. Centre for Advanced Biomedical Imaging, Paul O'Gorman Building, University College London, 72 Huntley Street, London WC1E 6DD, UK. Fax: +44 20 7905 2358. E-mail address: [email protected] (M.F. Lythgoe). 1 These authors contributed equally to this manuscript. 0014-4886/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.expneurol.2010.06.015
compensation phase occurs within the first 30 min of a seizure and is characterised by recruitment of physiological mechanisms such as increased blood flow and alterations in pH and glucose metabolism (Lothman and Collins, 1981; Meldrum and Nilsson, 1976). Subsequently, the brain enters the decompensation phase during which several physiological processes, such as cerebral autoregulation, fail and neuronal death may begin to occur (Meldrum and Nilsson, 1976). Since autoregulation maintains local blood flow to the tissue, it has been hypothesised that a failure may result in local ischaemia that could exacerbate injury (Meldrum and Nilsson, 1976). It is possible that such a local ischaemia is part of the pathophysiological process that underlies the vulnerability of the hippocampus to status epilepticus induced injury. The ability to non-invasively image cerebral haemodynamics during seizures has been hampered by a lack of methods that facilitate the mapping of CBF with high temporal and spatial resolution, over a wide range of blood flows (Calamante et al., 1999). Arterial spin labelling (ASL) is an MRI technique for measuring CBF that uses magnetically labelled blood water as an endogenous tracer (Thomas et al., 2000). This technique does not have a limitation on the number of repeat measurements in a single study, as it does not
M. Choy et al. / Experimental Neurology 225 (2010) 196–201
require the injection of an exogenous tracer. This makes ASL well suited for the continuous monitoring of the timecourse of CBF over durations of minutes to hours, e.g. to monitor the heterogeneous evolution of tissue perfusion after experimental ischemia and reperfusion in animal models (Lythgoe et al., 2002). Therefore, it is ideal for evaluating the timecourse of cerebral blood flow changes associated with status epilepticus. The images generated with ASL can be converted into CBF maps (ml/100 g/min) and are particularly suited to measuring quantitative regional blood flow changes over the course of an epileptic seizure. Despite these benefits, ASL has yet to be used to investigate seizures directly although it has been used for investigating the pathology following seizures (Choy et al., 2010; Ndode-Ekane et al., 2010). Only a few studies have imaged the brain during seizure activity with MRI. Magnetic resonance diffusion-weighted imaging (DWI) has been shown to be sensitive we investigate whether to regional changes during seizures and most likely reflects microstructural changes in the tissue (van Eijsden et al., 2004; Engelhorn et al., 2007; Zhong et al., 1995). A recent study used BOLD fMRI to demonstrate that seizure activity in rats was associated with negative BOLD signals in hippocampus, indicating a regional hemodynamic change (Schridde et al., 2008). Perfusion MRI has been used in a previous study, which suggested that there is a relationship between the maximum decrease in perfusion and subsequent neuronal loss using a gadolinium-based contrast agent (Engelhorn et al., 2005). However, these techniques do not allow the characterisation of a timecourse and therefore regional blood flow changes as a function of time during the course of status epilepticus. In this study, we hypothesise that the brain haemodynamic responses to seizures change over time, with the greatest response occurring during the first 30 min of a seizure. We also hypothesise that the haemodynamic response of the hippocampus to seizures is impaired when compared to that of the cerebral cortex. Therefore, the aim of this study was to use continuous arterial spin labelling (CASL) in the pilocarpine model of status epilepticus in rats to characterise the regional blood flow changes from before the onset of SE to seizure termination.
197
rats (300–350 g) were randomly divided into two groups and anaesthetised with either isoflurane (1.5% in a 60/40 N2O:O2 mix delivered at 1 l/min) (n = 6) or fentanyl citrate (300 μg/kg, i.p) and medetomidine (300 μg/kg, i.p) (n = 7). Tungsten wire electrodes (50 μm diameter, Science Products GMBH, Germany) were implanted bilaterally at symmetrical points in the hippocampus for EEG recording (AP, −4 mm from bregma; ML, 2 mm, and DV, 3.2 mm from the neocortex). The hippocampal electrodes were referenced to a ground electrode attached to the tail of the animal. EEG was acquired continuously until the end of the experiment. Methylscopolamine (1 mg/kg, i.p., Sigma-Aldrich) was injected i. p. to reduce the peripheral cholinergic effects of pilocarpine, followed 15–20 min later by pilocarpine hydrochloride injection (375 mg/kg, i. p., Sigma-Aldrich). If seizure activity was detected on EEG, then 90 min after the onset of continuous seizure activity, diazepam (10 mg/kg, i.p., Phoenix Pharma Ltd, Gloucester, UK) was given to attenuate the seizure. Pilocarpine hydrochloride and methylscopolamine were freshly prepared prior to administration in 0.9% saline. Reflex behaviour was used to assess the level of anaesthesia. Two reflex behaviours were used to provide an estimation of the depth of anaesthesia required for surgical procedures: paw pinch reflex and corneal reflex (Alves et al., 2009; Green, 1982). The depth of anaesthesia was tested every 10 min using these reflex behaviours to determine whether adequate anaesthesia was maintained throughout the experiment. If the animal retained either reflex during any period of the protocol then, for fentanyl, an additional dose of 100 μg/kg of was given, or for isoflurane, the inspired dose was increased. The presence of status epilepticus was investigated using EEG which was recorded continuously using a Bioamp differential amplifier interfaced with a Powerlab data acquisition system (AD Instruments, Australia). Signals were filtered between 1 and 120 Hz, a 50-Hz notch filter was used to reduce noise, with a sampling rate of 1024 samples per second. Analysis of electrophysiological data was carried out off-line on a Pentium computer, using Chart software (AD Instruments, Australia).
Methods All animal care and procedures were carried out in accordance with the UK Animals (Scientific Procedures) 1986 Act. As MR imaging of small animals typically requires the use of anaesthesia, an initial study was performed to determine whether pilocarpine can be used to induce status epilepticus under anaesthesia. Following these initial experiments, we investigated the regional nature of CBF changes during pilocarpine-induced SE using ASL. Study 1. Pilocarpine-induced SE under anaesthesia Anaesthesia has been shown to be effective for abolishing seizure activity, and used clinically in seizures refractory to conventional anticonvulsant therapy (Shorvon, 1994). However, it is commonly required in MR imaging of small animals as anaesthesia reduces stress and limits motion, which can lead to artefacts in the acquired images (Lukasik and Gillies, 2003). It was therefore necessary to determine that seizures could be induced with pilocarpine under anaesthesia. Isoflurane is a commonly used gaseous anaesthetic for MR imaging, due to, in part, its rapid clearance and ease of management for the animal in the MR system (Lukasik and Gillies, 2003), although it is known to be an effective anti-convulsant (Shorvon, 1994). Fentanyl, by itself, has been shown to be an effective method for sedation and subsequent induction of seizures (Engelhorn et al., 2005), but does not meet the requirement for anaesthesia (Flecknell, 2009). However, fentanyl in combination with medetomidine, can provide effective anaesthesia (Flecknell, 2009). Therefore we investigated the ability of pilocarpine to induce SE under these two anaesthetic regimens. Thirteen adult male Sprague–Dawley
Study 2. CBF measurements with MRI 19 adult male Sprague–Dawley rats (300–400 g, Charles Rivers) were divided randomly into two groups: pilocarpine (n = 13 of which 6 were discarded because of movement artefacts in the images) or saline (n = 6). Once the animals had been prepared for imaging (see below), scopolamine (1 mg/kg) was administered. T1-weighted images were acquired at the beginning of the experiment for the purpose of CBF quantification. Thereafter ASL was acquired continuously for the duration of the experiment. The experimental design is outlined in Fig. 1. The first 30 min of MRI were acquired for baseline measurements after which the animals were administered either pilocarpine (375 mg/kg) or saline. Bench experiments (see study 1) indicated that SE began approximately 30 min following pilocarpine administration; therefore in order to obtain data from 2 h of continuous seizure activity the animals were imaged for 150 min before diazepam (10 mg/kg) was given. After diazepam the rats were imaged for up to a further 30 min. The animals were visually assessed for seizure activity following each set of perfusion acquisitions. Animals were anaesthetised with intraperitoneal injections of fentanyl citrate (300 μg/kg) and medetomidine (300 μg/kg). Additional doses of fentanyl (100 μg/kg) were given if the paw pinch reflex was present in order to maintain anaesthesia. O2 was delivered continuously via a nose cone at a rate of 1 l/min. Rats were placed on a speciallydesigned animal holder to minimize motion artifacts. Physiological monitoring included electrocardiography (ECG) recordings and rectal temperature recordings. Temperature was maintained at 37 °C± 2.
198
M. Choy et al. / Experimental Neurology 225 (2010) 196–201
Fig. 1. Experimental design for MRI-ASL measurements during SE. The key events that occur over the course of MR imaging are as follows: 30 min of baseline acquisition before pilocarpine administration at time 0, SE onset occurs approximately 30 min after administration and diazepam was given 120 min after pilocarpine.
Images were acquired using a volume transmitter coil and a separate decoupled surface receiver coil. A central coronal imaging slice was selected that included the hippocampus (−4.8 mm from bregma). The imaging slice was determined by visual inspection of anatomical landmarks. All sequences were acquired from a 2-mmthick coronal slice. A low average perfusion acquisition was used in this study to reduce the effects of motion. For non-invasive CBF measurement, we used the continuous arterial spin labelling (CASL) method (Alsop and Detre, 1996), based on spin-echo echo-planar imaging (EPI) with interleaved adiabatic fast-passage inversion and control measurements. The centre of the inversion plane was situated 2 mm behind the cerebellum in all animals (approximately 10 kHz frequency offset); for the control image the same offset value was used, but with the sign reversed. The CASL labelling pulse duration was 3 s, and a post-labelling delay time of 500 ms was used to minimise transit time effects and intravascular artefacts (Thomas et al., 2006). The echo time was 36 ms, the interexperiment delay was 1000 ms, and 22 averages were acquired. A non-slice-selective inversion recovery (IR) EPI sequence was used to obtain the tissue T1 and the spin density (M0) parameters necessary for subsequent CBF quantification. IR-EPI images were acquired with: TR = 2000 ms, TE = 36 ms, an inversion time (TI) array of 284 ms, 434 ms, 634 ms, 834 ms, 1234 ms, 1734 ms, 2734 ms, and 3734 ms, and 22 averages were used. The IR data were fitted to obtain T1, α0 and M0 assuming a mono-exponential recovery function. Subsequently, these values were used to calculate CBF from the difference between the labelled and control CASL images (Alsop and Detre, 1996). The following assumptions were used to determine CBF: T1 of arterial
blood = 1.5 s, blood:brain partition coefficient for water (λ) = 0.9, efficiency of the spin labelling pulse = 0.71, tissue transit time = 250 ms, arterial transit time = 200 ms. Durations for each acquisition were: T1 = 11 min and CASL = 4 min. Each animal was scanned for a total time of 3 h. T1 maps were reconstructed using in-house software written in Matlab version 6.5 (MathWorks, Massachusetts, USA). As motion artefacts may lead to CBF quantification errors, perfusion-weighted images were visually inspected, and images with visible motion artefacts were excluded from the study. However, removal of these images decreased the signal to noise of these images, and therefore limited the precision of CBF measurements. We therefore averaged the images across 5 distinct periods to achieve reasonable precision for CBF, which were determined by the EEG experiments described in study 1 and from previous studies reviewed by Lothman (Lothman, 1990). These were: 1. baseline. 2. Within the first 30 min following pilocarpine administration. This is the period before SE occurs. 3. 30–60 min after pilocarpine. From our previous experiments, this period coincides with the first 30 min of SE and has been associated with the compensation phase of SE. 4. 60–150 min after pilocarpine. This period is associated with the latter stages of SE and this is considered to be the decompensation phase. 5. After approximately 150 min, diazepam was administered to the animal, and this period should reflect the gradual cessation of epileptic activity.
Fig. 2. Representative intrahippocampal EEG recording in a fentanyl/medetomidine anaesthetised rat. (a) 3 h EEG acquisition for the duration of experiment; (b) baseline EEG; (c) EEG desynchronisation post-pilocarpine (375 mg/kg); (d) increase in EEG indicating status epilepticus; (e) decrease in EEG following diazepam (10 mg/kg). Black arrow denotes pilocarpine administration, open arrow indicates diazepam administration. Data presented as mean ± SEM.
M. Choy et al. / Experimental Neurology 225 (2010) 196–201
199
Table 1 Mean regional CBF measurements before, during, and after pilocarpine. Regions include the parietal cortex, hippocampus and the thalamus. Data presented as mean ± SEM. CBF (ml/100 g/min)
Parietal cortex
Hippocampus
Thalamus
Baseline 0–30 min 30–60 min 60–120 min 120–150 min
122 ± 11 234 ± 32 121 ± 24 132 ± 32 165 ± 24
96 ± 14 118 ± 21 65 ± 19 105 ± 15 90 ± 27
109 ± 15 155 ± 24 108 ± 12 153 ± 30 150 ± 28
Three regions of interest (parietal cortex, hippocampus and thalamus) were delineated on all processed images (Fig. 3) and analysed with inhouse Matlab software and ImageJ version 1.38x (National Institutes of Health, USA). Statistical analysis For CBF measurements, a 3-way repeated measures ANOVA was conducted with contrasts to the baseline scans and the hippocampus. The main effects were for group (pilocarpine, saline), time (baseline, pilocarpine, SE, SE late, diazepam) and anatomy (parietal cortex, hippocampus, thalamus). All data presented as mean ± standard error of the mean (sem).
Results
Fig. 4. Ratios of parietal cortex CBF to hippocampal CBF over the course of SE. The periods are divided up as follows: 0–30 min is the period following pilocarpine injection; 30– 120 min during SE and; 120–150 min is following the injection of diazepam. Dashed line with open diamonds is the saline treated rats (n = 6), and full line with filled squares is the pilocarpine treated rats (n= 7). Data presented as mean± SEM, and significant statistical differences are indicated by * (p b 0.05).
previous reports in this model without anaesthesia (83%) (Mello et al., 1993) (Fig. 2d). Of the animals that did progress onto SE, the mean SE onset time following pilocarpine administration was 22.8 ± 1.9 min. In these animals, seizure activity persisted until the administration of diazepam, at which point the amplitude of the EEG gradually reduced (Fig. 2e).
Induction of SE under anaesthesia CBF measurements Throughout the duration of the experiments no reflex behaviours were observed using either anaesthetic regime. Following isofluraneanaesthesia, animals did not show any sign of seizure behaviour following pilocarpine. However, the fentanyl/medetomidine anaesthetised animals displayed facial automatisms and tonic extensions of the forelimbs and hindlimbs, in which the entire body of the rat became outstretched. These tonic extensions lasted for a few seconds, occurred intermittently, and only during the period of increased EEG activity. Fentanyl/medetomidine animals (6/7) displayed all of these behavioural responses. These effects are the main reason for the movement artefact that we removed from our data. Both groups of anaesthetised animals displayed a reduction in EEG amplitude following pilocarpine administration. The reduced amplitude persisted for the duration of the experiments in the isofluraneanesthetised animals. In the fentanyl/medetomidine-anaesthetised animals (Fig. 2), the reduced amplitude continued until the onset of SE. None of the animals under isoflurane anaesthesia progressed to SE (n = 6). In contrast, 6 out of 7 (85%) animals under fentanyl/ medetomidine anaesthesia progressed to SE, which is similar to
In total, 7 animals were included for analysis from the pilocarpine group (Table 1). In the control group, the estimated CBF remained relatively stable following saline injection (n = 6). Following pilocarpine injection, we observed an increase in parietal cortex blood flow relative to hippocampal blood flow at 0–30 min (Fig. 3.) (F = 14.9, p b 0.01) and at 30–60 min (F = 7.4, p = 0.02), but not after that time. The cortical blood flow was 2.84 ± 0.5 times more than hippocampal blood flow 30–60 min (Fig. 4). However, in the saline control group the cortical blood flow was only 1.46 ± 0.2 times more than hippocampal blood flow 30–60 min (F = 6.87, p = 0.03) Furthermore the hippocampus was the only region in which the mean blood flow decreased below the mean baseline level (baseline = 96 ± 14 ml/min/100 g; 30–60 min = 65 ± 19 ml/min/100 g). No significant differences were found between parietal cortex and thalamus or hippocampus and thalamus. These data indicate hippocampus and the parietal cortex respond to pilocarpine in a different manner and that the hippocampus is hypoperfused compared to the parietal cortex (Table 2).
Fig. 3. Representative mean perfusion-weighted images during SE from (a) a saline-injected and (c) a pilocarpine-injected animal. Panel b shows a T1 map use to calculate quantitative blood flow and the ROIs are outlined. The regions are: (1) parietal cortex; (2) hippocampus; (3) thalamus. Note the limited perfusion change in the hippocampus compared to the cortex. Data presented as mean ± SEM.
200
M. Choy et al. / Experimental Neurology 225 (2010) 196–201
Table 2 Regional and time-dependent CBF differences before, during, and after pilocarpine. Note that significant differences were found between the hippocampus and the parietal cortex at 0–30 and 30–60 min after pilocarpine injection when compared to baseline.
Thalamus vs. Hippocampus
Parietal cortex vs. Hippocampus
Time period contrast with baseline (min)
F
p
0–30 30–60 60–120 120–150 0–30 30–60 60–120 120–150
0.39 0.50 0.07 0.07 14.92 7.35 3.05 0.72
0.54 0.49 0.80 0.80 b 0.01 0.02 0.11 0.42
Discussion There were two main findings from this ASL-MRI study. Firstly we show that pilocarpine can be used to induce SE under fentanyl/ medetomidine anaesthesia, and that seizure activity occurs in the hippocampus. Secondly, during the course of pilocarpine-induced SE, cortical perfusion is high relative to hippocampal perfusion. An important component of the current study is the development of a model in which seizure activity could be induced without the behavioural manifestations of SE, such as the convulsive movement, confounding the experiments. We were unable to achieve this using isoflurane as an anaesthetic regime as no seizure activity was observed. As fentanyl with medetomidine has previously been shown to elicit anaesthesia suitable for surgical procedures in rodents (Meert and De, 1994; Wolfensohn and Lloyd, 1998) we investigated this anaesthetic protocol and confirmed that pilocarpine could be used to induce SE under this regime and therefore we were able to investigate cerebral blood flow associated with SE. Whether fentanyl/medetomidine anaesthesia has an effect on the elicited seizure remains unclear. The classes of drugs that fentanyl and medetomodine fall under, opiates and α2-adrenergic agonists, respectively, have been shown to affect seizure activity. Opiates have been found to potentiate seizures (Cherng and Wong, 2005), and α2adrenergic agonists have been shown to reduce seizure threshold (Mirski et al., 1994) as well as suppress seizure activity (Halonen et al., 1995). However, the interaction between these two classes of drug on seizures has not been well documented. Comparing our data with nonanaesthetised pilocarpine-induced seizures, we observed similar seizure onset times as well as a similar number of animals that progress on to SE (Cavalheiro et al., 2006). Also we demonstrate continued hippocampal seizure activity for at least 90 min, and that the seizure activity was attenuated following diazepam administration, indicating similarities between the pilocarpine models. Despite these similarities, both anaesthesia and pilocarpine affect central neurotransmitter systems and at some level induced seizures may be affected by anaesthesia, which will require further investigation. However, for the purposes of this study of investigating the time-dependent and regional CBF changes, the primary goal was eliciting prolonged hippocampal seizure activity, which was achievable in anaesthetised animals. Previous studies that have investigated the regional CBF profile during SE have typically used autoradiography techniques, which can provide regional CBF measurements that are limited to a single time point, or electrochemical techniques that provide continuous CBF measurements at a single small anatomical point in the brain (Meldrum and Nilsson, 1976; Shih and Scremin, 1992; Tanaka et al., 1990; Kreisman et al., 1991; Andre et al., 2002; Nersesyan et al., 2004). Therefore data providing whole regional CBF changes over time during SE are still required. We used ASL-MRI to investigate the CBF time course before, during and after pilocarpine-induced SE across a number of brain structures. Repeated ASL measurements demonstrated an initial CBF increase following pilocarpine administration in the three regions investigated.
The data provide good evidence for a reduced haemodynamic response in the hippocampus when compared to the cortex. A similar pattern was also observed in the thalamus relative to the hippocampus although the degree of perfusion change in the cortex tended to be greater than the thalamus. Absolute values of CBF indicated that the flow was not constant during the seizure process, which may represent loss of cerebral autoregulation after 30 min of seizure activity (Lothman, 1990). Our hippocampal EEG measurements indicate that seizure activity occurs continuously in the hippocampus under our anaesthetic regimen and persists until diazepam administration; therefore seizure activity alone is unlikely to account for these observed CBF variations. One drawback in our study was the time resolution, which limits a full assessment of the hemodynamic changes, and a further characterisation is necessary to elucidate the precise nature of prolonged seizure activity and the absolute CBF time-dependent changes. Our data demonstrate a region-specific response during seizure activity. The changes in parietal cortical CBF are consistent with previous studies during the periods of seizure activity. Accompanying increases in glucose utilisation and CMRO2 have also been reported (Shih and Scremin, 1992; Andre et al., 2002; Pinard et al., 1984; Tanaka et al., 1990; Kreisman et al., 1991). These data suggest that increases in CBF act to compensate for the increase in nutrient demand that the activated neurons require in the early phase of seizures. We observed that the hippocampus was hypoperfused compared to the cortex, which is endorsed by previous studies investigating bicuculline-induced seizures, where the authors have show a negative BOLD response in the hippocampus which they attributed to a mismatch between CMRO2 and CBF (Schridde et al., 2008; DeSalvo et al., 2010). An autoradiography study using a kainic acid seizure model also reported a limited hippocampal CBF response that was insufficient for the increase in metabolic demand (Tanaka et al., 1990), which has also been observed in the soman model of SE (Shih and Scremin, 1992). Taken together, these data suggest that the blood flow to the hippocampus may not meet local increases in energy demand during seizure activity, leading to a regional ischaemia. This may reflect the later period of SE for which there is a hypothesised failure of cerebral autoregulation, and may explain the reduction in blood flow we observed. Nevertheless contradictory reports exist. Substantial increases in hippocampal flow relative to the cortex have also been reported in the kainic acid model using helium clearance, an effect not observed in the bicuculline model (Pinard et al., 1987). This may suggest that the method of seizure induction may have an effect on the haemodynamic response to prolonged seizures. Whether the method of seizure induction with anaesthesia has or has not influenced the haemodynamic response is unclear, however, the non-anaesthetised soman model used by Shih and Scremin acts, like pilocarpine, to enhance cholinergic activity also reported a blunted CBF response, suggesting that this may be a property of the induced seizures (Shih and Scremin, 1992). These data therefore suggest that a relative hypoperfusion may contribute to injury during SE, which may have implications for treatment of SE in humans. In conclusion, this study indicates that regional CBF changes occur in vivo during pilocarpine-induced SE under fentanyl/medetomidine anaesthesia. When compared to cortical perfusion, the hippocampus CBF was consistently lower and these data support a hypothesis that this reduced blood flow may play a role in the selective vulnerability of the hippocampus to SE, which may, at least in part, underlie the seizureinduced hippocampal damage observed in animals and humans. Acknowledgments This work was supported by Epilepsy Research UK, Wellcome Trust, the Biotechnology and Biological Sciences Research Council and the British Heart Foundation. The authors thank Dr. Martin King for his advice on statistics, and Prof. Matthew Walker for his guidance on the pilocarpine model.
M. Choy et al. / Experimental Neurology 225 (2010) 196–201
References Alsop, D.C., Detre, J.A., 1996. Reduced transit-time sensitivity in noninvasive magnetic resonance imaging of human cerebral blood flow. J. Cereb. Blood Flow Metab. 16, 1236–1249. Alves, H.C., Valentim, A.M., Olsson, I.A., Antunes, L.M., 2009. Intraperitoneal anaesthesia with propofol, medetomidine and fentanyl in mice. Lab Anim. 43, 27–33. Andre, V., Henry, D., Nehlig, A., 2002. Dynamic variations of local cerebral blood flow in maximal electroshock seizures in the rat. Epilepsia 43, 1120–1128. Calamante, F., Thomas, D.L., Pell, G.S., Wiersma, J., Turner, R., 1999. Measuring cerebral blood flow using magnetic resonance imaging techniques. J. Cereb. Blood Flow Metab. 19, 701–735. Cavalheiro, E.A., Naffah-Mazzacoratti, M.G., Mello, L.E., Leite, J.P., 2006. Chapter 35, Models of Seizures and Epilepsy. Elsevier Inc. Cherng, C.H., Wong, C.S., 2005. Effect of fentanyl on lidocaine-induced convulsions in mice. Pharmacology 75, 1–4. Choy, M., Cheung, K.K., Thomas, D.L., Gadian, D.G., Lythgoe, M.F., Scott, R.C., 2010. Quantitative MRI predicts status epilepticus-induced hippocampal injury in the rat lithium-pilocarpine model. Epilepsy Res. 88 (2–3), 221–230. Desalvo, M.N., Schridde, U., Mishra, A.M., Motelow, J.E., Purcaro, M.J., Danielson, N., Bai, X., Hyder, F., Blumenfeld, H., 2010. Focal BOLD fMRI changes in bicuculline-induced tonic–clonic seizures in the rat. Neuroimage 50 (3), 902–909. Engelhorn, T., Doerfler, A., Weise, J., Baehr, M., Forsting, M., Hufnagel, A., 2005. Cerebral perfusion alterations during the acute phase of experimental generalized status epilepticus: prediction of survival by using perfusion-weighted MR imaging and histopathology. AJNR Am. J. Neuroradiol. 26, 1563–1570. Engelhorn, T., Weise, J., Hammen, T., Bluemcke, I., Hufnagel, A., Doerfler, A., 2007. Early diffusion-weighted MRI predicts regional neuronal damage in generalized status epilepticus in rats treated with diazepam. Neurosci. Lett. 417, 275–280. 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. Flecknell, P., 2009. Laboratory animal anaesthesia, Third Edition. Elsevier Inc. Green, C.J., 1982. Animal anaesthesia. Laboratory Animals Ltd. Halonen, T., Kotti, T., Tuunanen, J., Toppinen, A., Miettinen, R., Riekkinen, P.J., 1995. Alpha 2-adrenoceptor agonist, dexmedetomidine, protects against kainic acidinduced convulsions and neuronal damage. Brain Res. 693, 217–224. Kreisman, N.R., Magee, J.C., Brizzee, B.L., 1991. Relative hypoperfusion in rat cerebral cortex during recurrent seizures. J. Cereb. Blood Flow Metab. 11, 77–87. Lothman, E., 1990. The biochemical basis and pathophysiology of status epilepticus. Neurology 40, 13–23. Lothman, E.W., Collins, R.C., 1981. Kainic acid induced limbic seizures: metabolic, behavioral, electroencephalographic and neuropathological correlates. Brain Res. 218, 299–318. Lukasik, V.M., Gillies, R.J., 2003. Animal anaesthesia for in vivo magnetic resonance. NMR Biomed. 16, 459–467. Lythgoe, M.F., Thomas, D.L., Calamante, F., 2002. MRI measurement of cerebral perfusion and the application to experimental neuroscience. In: van-Bruggen, N., Roberts, T.P. (Eds.), Biomedical Imaging in Experimental Neuroscience. CRC Press, Boca Raton, Florida. Meldrum, B.S., Nilsson, B., 1976. Cerebral blood flow and metabolic rate early and late in prolonged epileptic seizures induced in rats by bicuculline. Brain 99, 523–542.
201
Meert, T.F., De, K.M., 1994. Potentiation of the analgesic properties of fentanyllike opioids with alpha 2-adrenoceptor agonists in rats. Anesthesiology 81, 677–688. Mello, L.E., Cavalheiro, E.A., Tan, A.M., Kupfer, W.R., Pretorius, J.K., Babb, T.L., Finch, D.M., 1993. Circuit mechanisms of seizures in the pilocarpine model of chronic epilepsy: cell loss and mossy fiber sprouting. Epilepsia 34 (6), 985–995. Mirski, M.A., Rossell, L.A., McPherson, R.W., Traystman, R.J., 1994. Dexmedetomidine decreases seizure threshold in a rat model of experimental generalised epilepsy. Anesthesiology 81, 1422–1428. Ndode-Ekane, X.E., Hayward, N., Gröhn, O., Pitkänen, A., 2010. Vascular changes in epilepsy: functional consequences and association with network plasticity in pilocarpine-induced experimental epilepsy. Neuroscience 166 (1), 312–332. Nersesyan, H., Herman, P., Erdogan, E., Hyder, F., Blumenfeld, H., 2004. Relative changes in cerebral blood flow and neuronal activity in local microdomains during generalized seizures. J. Cereb. Blood Flow Metab. 24, 1057–1068. Pinard, E., Tremblay, E., Ben-Ari, Y., Seylaz, J., 1984. Blood flow compensates oxygen demand in the vulnerable CA3 region of the hippocampus during kainate-induced seizures. Neuroscience 13, 1039–1049. Pinard, E., Rigaud, A.S., Riche, D., Naquet, R., Seylaz, J., 1987. Continuous determination of the cerebrovascular changes induced by bicuculline and kainic acid in unanaesthetized spontaneously breathing rats. Neuroscience 23, 943–952. Rona, S., Rosenow, F., Arnold, S., Carreno, M., Diehl, B., Ebner, A., Fritsch, B., Hamer, H.M., Holthausen, H., Knake, S., Kruse, B., Noachtar, S., Pieper, T., Tuxhorn, I., Luders, H.O., 2005. A semiological classification of status epilepticus. Epileptic Disord. 7, 5–12. Schridde, U., Khubchandani, M., Motelow, J.E., Sanganahalli, B.G., Hyder, F., Blumenfeld, H., 2008. Negative BOLD with large increases in neuronal activity. Cereb. Cortex 18, 1814–1827. Scott, R.C., King, M.D., Gadian, D.G., Neville, B.G., Connelly, A., 2006. Prolonged febrile seizures are associated with hippocampal vasogenic edema and developmental changes. Epilepsia 47, 1493–1498. Shih, T.M., Scremin, O.U., 1992. Cerebral blood flow and metabolism in soman-induced convulsions. Brain Res. Bull. 28, 735–742. Shorvon, S., 1994. Status Epilepticus. Cambridge University Press. Siesjo, B.K., Wieloch, T., 1986. Epileptic brain damage: pathophysiology and neurochemical pathology. Adv. Neurol. 44, 813–847 813-847. Tanaka, S., Sako, K., Tanaka, T., Nishihara, I., Yonemasu, Y., 1990. Uncoupling of local blood flow and metabolism in the hippocampal CA3 in kainic acid-induced limbic seizure status. Neuroscience 36, 339–348. Thomas, D.L., Lythgoe, M.F., Calamante, F., Pell, G.S., Ordidge, R.J., 2000. The measurement of diffusion and perfusion in biological systems using magnetic resonance imaging. Phys. Med. Biol. 45, R97–R138. Thomas, D.L., Lythgoe, M.F., van der Weerd, L., Ordidge, R.J., Gadian, D.G., 2006. Regional variation in cerebral blood flow and arterial transit time in the normal and hypoperfused rat brain measured using continuous arterial spin labeling MRI. J. Cereb. Blood Flow Metab. 26, 274–282. van Eijsden, P., Notenboom, R.G., Wu, O., de Graan, P.N., van Nieuwenhuizen, O., Nicolay, K., Braun, K.P., 2004. In vivo (1)H magnetic resonance spectroscopy, T(2)weighted and diffusion-weighted MRI during lithium-pilocarpine-induced status epilepticus in the rat. Brain Res. 1030, 11–18. Wolfensohn, S., Lloyd, M., 1998. Handbook of laboratory animal management and welfare, Second Ed. Blackwell Publishing. Zhong, J., Petroff, O.A., Prichard, J.W., Gore, J.C., 1995. Barbiturate-reversible reduction of water diffusion coefficient in flurothyl-induced status epilepticus in rats. Magn. Reson. Med. 33, 253–256.