In vivo imaging of seizure activity in a novel developmental seizure model

Abstract

The immature brain is exceptionally susceptible to seizures. However, it remains unclear whether seizures occurring during development affect critical processes underlying neural circuit formation, leading to long-term functional consequences. Here we characterize a novel in vivo model system of developmental seizures based on the transparent albino Xenopus laevis tadpole, which allows direct examination of seizure activity, and seizure-induced effects on neuronal development within the intact unanesthetized brain. Pentylenetetrazol (PTZ), kainic acid, bicuculline, picrotoxin, 4-aminopyridine, and pilocarpine were tested for their ability to induce behavioral seizures in freely swimming tadpoles when bath applied. All six chemoconvulsants consistently induced similar patterns of abnormal behavior in a dose-dependent manner, characterized by convulsive clonus-like motor patterns with periods of behavioral arrest. Extracellular field recordings demonstrated rhythmic synchronous epileptiform electrographic responses induced by convulsants irrespective of mechanism of action, that could be terminated by the anti-epileptic drug valproate. PTZ-induced seizures were further characterized using in vivo two-photon fluorescence imaging of neuronal calcium dynamics, in unanesthetized immobilized tadpoles. Imaging of calcium dynamics during PTZ-induced seizures revealed waves of neural activity propagating through large populations of neurons within the brain. Analysis of single-cell responses demonstrated distinct synchronized high-amplitude calcium spikes not observed under baseline conditions. Similar to other developmental seizure models, prolonged seizures failed to induce marked neuronal death within the brain, detected by cellular propidium iodide incorporation in vivo or TUNEL labeling. This novel developmental seizure model system has distinct advantages for controlled seizure induction, and direct visualization of both seizure activity and seizure-induced effects on individual developing neurons within the intact unanesthetized brain. Such a system is necessary to address important questions relating to the long-term impact of common perinatal seizures on developing neural circuits.

Introduction

The highest incidence of seizures occurs during the first years of life, and in patients who develop epilepsy recurrent seizures often begin during childhood (Cowan, 2002, Hauser, 1995). Early brain development is marked by heightened neural plasticity and dynamic neural circuit formation involving ongoing activity-dependent neuronal growth, synaptogenesis and circuit refinement (Katz and Shatz, 1996, Spitzer, 2006). While these ongoing maturational processes may underlie the immature brain's relatively low seizure threshold and promote seizure expression (Haas et al., 1990), it remains unclear whether seizures, in turn, interfere with activity-dependent maturation of neural circuits.

Considerable debate has focused on whether seizures induce alterations in brain circuit development leading to persistent neural dysfunction or to an increased susceptibility to psychiatric or neurological disorders later in life (Ben-Ari and Holmes, 2006, Swann, 2004). While some clinical studies examining the long-term consequences of early-life seizures have reported an increased risk for the subsequent development of epilepsy, reduced cognitive function, or psychiatric illness (Cornaggia et al., 2006, Gaitatzis et al., 2004, Vestergaard et al., 2005, Vestergaard et al., 2007), others have found no such link (Scher, 1997, Shinnar and Hauser, 2002, Verity et al., 1998). Unfortunately, clinical studies are confounded by the inherent variability of patient populations, such as variation in number, type and frequency of seizures, age of seizure onset, individual genetic susceptibilities, underlying etiology, medication, and other environmental influences (Holmes, 1991, Scher, 1997). Furthermore, when seizures precede the onset of neurological dysfunction, it is difficult to determine whether seizures directly induce the observed dysfunction or whether both arise from shared pathological origins (Tsopelas et al., 2001). It is also possible that gross measures currently employed to examine the long-term effects of seizures may be unable to detect subtle changes in developing neural circuits.

Studies utilizing experimental animal models have been able to address these problems by offering increased control over experimental variables and more precise measures of seizure induced changes. In immature rodent models, animals experiencing early-life seizures demonstrate a remarkable resistance to seizure-induced neuronal death and reactive axonogenesis commonly seen in adult animals (Haas et al., 2001b, Sperber et al., 1991). However, developmental seizures have been shown to induce permanent changes in behavior, learning and memory, and an increased susceptibility to seizures in adulthood (Holmes et al., 1988, Lynch et al., 2000). Given that experimentally induced seizures in immature animals permanently alter neural function, it is of critical importance to identify the precise molecular and structural alterations that underlie these changes.

Recent advances in imaging methodologies coupled with improved fluorescent indicators of intracellular events has enabled the direct in vivo visualization of neuronal activity and morphological growth occurring during neuronal development (Garaschuk et al., 2006a, Niell and Smith, 2004, Sanchez et al., 2006, Sin et al., 2002, Haas et al., 2001a). However, the application of these techniques for the study of seizure-induced effects on neural development and circuit formation using conventional mammalian developmental seizure models is limited due to: (i) the requirement of anesthetic agents during in vivo imaging of the brain during seizures, which disrupt and reduce neuronal excitability (Ishizawa, 2007); (ii) the invasive surgical procedures required are impractical during periods of early brain development (Mizrahi et al., 2004, Rensing et al., 2005, Zeng et al., 2007). An alternative approach to overcome these limitations is the use of non-mammalian model organisms that are already in widespread use in neurodevelopment research (Baraban, 2007).

Here we describe a novel developmental seizure model system specifically designed to address questions of how seizures are initiated and propagate through immature brain circuits, and how seizures affect subtle yet critical developmental processes including neuronal growth, synaptogenesis and brain circuit formation. This model system is based on the transparent albino Xenopus laevis tadpole, a well-studied vertebrate used extensively for in vivo imaging of neuronal growth and synaptogenesis (Cantallops et al., 2000, Haas et al., 2006, Sanchez et al., 2006, Sin et al., 2002). The transparency of this organism allows direct in vivo imaging of the growth and activity of individual or large populations of neurons within the intact developing brain. Furthermore, immobilization of tadpoles by reversible paralytics and immersion in agar circumvents the need for anesthetic agents during electrophysiological or imaging experiments. We have characterized experimentally induced seizures in tadpoles using behavioral assessment, measures of cell death, and in vivo examination of neural activity during seizures, using electrophysiological recordings and imaging of intracellular calcium dynamics within the intact unanesthetized brain.