A revised Racine's scale for PTZ-induced seizures in rats

Abstract

Behavioral scoring is commonly used to access seizure intensity in different seizure models. Racine's scale, originally developed for the amygdala-kindling model, is also frequently used as an intensity measurement in other experimental seizure or epilepsy models. The aim of the present study is to assess the validity of Racine's scale as an adequate seizure intensity measurement for the Pentylenetetrazole (PTZ) model. Male adult Wistar rats received systemic injections of PTZ starting with an initial dose of 20 mg/kg added up by 10 mg/kg every 15 min until the occurrence of a 5 minute lasting convulsive seizure. Simultaneous EEG and video recordings were made. The PTZ-induced seizures gradually increased in intensity. Eleven behavioral categories were identified and statistically analyzed. Six different seizure intensity categories were found to differ from each other based on differences in onset latencies, the pattern of occurrence during high or low doses of PTZ and the EEG pattern. These categories were quite different from those of Racine's scale.

We suggest that Racine's scale is not adequate for the assessment of the seizure intensity of PTZ-induced seizures and that an alternative scale with the six proposed behavioral seizure categories is a more adequate description of PTZ-induced seizures.

Introduction

Racine's scale (RS) is one of the most frequently used tools to assess the intensity of a seizure in rodent models of experimental epilepsy. RS categorizes five stages of intensity, and it is based on the behavioral repertoire of the animals during a seizure, including “mouth and facial movements” (intensity stage 1); “head nodding” (stage 2); “forelimb clonus” (stage 3); seizures characterized by rearing, (stage 4) and seizures characterized by rearing and falling (stage 5) [1].

Racine developed the scale by investigating the relationship between EEG-changes and the development of motor seizures in the amygdala-kindling model, characterized by partial seizures with secondary generalization [2], [3]. Nowadays, RS is frequently applied to many other models for seizures and epilepsy [e.g. [4], [5], [6]].

It is however questionable, whether the use of RS for the assessment of seizure intensities in other epilepsy or seizure models is justifiable, given the well known relation between activated brain part and corresponding expressed behavior [7].

Others have also questioned the use of this scale when rats were kindled in other limbic sites [8]. These authors demonstrated that rats kindled in the hippocampus developed stage 5-seizures without the exhibition of stage 1 and stage 2 responses. These stages were replaced by expressions of “wet-dog shakes” instead. In addition it was found that rats kindled from the perirhinal cortex rapidly display stage 3 to 5 seizures, while stage 1 to 2 seizures develop later on [8].

Even more questionable is the validity of RS for the use in models, which do not rely on kindling induced epileptogenesis from limbic structures and exhibit different brain activation patterns. The need for different intensity classification systems has been demonstrated for audiogenic seizures, which represent secondary generalized seizures originating from the brainstem: Jobe et al. identified nine different intensity stages in generalized epilepsy prone rats (GEPRs) and Garcia-Cairasco et al. developed an ethological cluster analysis to describe the progression of audiogenic seizures in Wistar audiogenic rats (WARs), that incorporates probability relationships between different behavioral expressions, demonstrating the complexity of behavioral expressions [9], [10].

Another widely used model, to which the RS is often applied to, is the Pentylenetetrazole (PTZ) model. Systemic injection of the GABA_A antagonist PTZ induces primary generalized seizures. In small doses, PTZ has been used as a model for absence seizures and in higher doses it produces convulsive seizures. It is, together with the Maximal Electroshock model (MES), one of the most frequently used models for antiepileptic drug assessment [3], [11]. Last but not least, the intensity scale is also often used to determine a humane endpoint of an experiment. For this reason, an inadequate description of the seizures might question whether it can be reliably be used for this function. The goal of the current study is to evaluate whether the RS is an adequate assessment tool for the intensity of PTZ-induced seizures. Furthermore, we want to investigate whether it is possible to establish a more adequate, statistical-based seizure intensity scale for the PTZ model.