Address correspondence to Norberto Garcia-Cairasco, PhD, Neurophysiology and Experimental Neuroethology Laboratory, Department of Physiology, Ribeirão Preto School of Medicine, Ribeirão Preto 14049-900, São Paulo, Brazil. E-mail: email@example.com
Purpose: The role of the superior colliculus (SC) in seizure expression is controversial and appears to be dependent upon the epilepsy model. This study shows the effect of disconnection between SC deep layers and adjacent tissues in the expression of acute and kindling seizures.
Methods: Subcollicular transections, ablation of SC superficial and deep layers, and ablation of only the cerebral cortex were evaluated in the Wistar audiogenic rat (WAR) strain during acute and kindled audiogenic seizures. The audiogenic seizure kindling protocol started 4 days after surgeries, with two acoustic stimuli per day for 10 days. Acute audiogenic seizures were evaluated by a categorized seizure severity midbrain index (cSI) and kindled seizures by a severity limbic index (LI).
Results: All subcollicular transections reaching the deep layers of the SC abolished audiogenic seizures or significantly decreased cSI. In the unlesioned kindled group, a reciprocal relationship between limbic and brainstem pattern of seizures was seen. The increased number of stimuli provoked an audiogenic kindling phenomenon. Ablation of the entire SC (ablation group) or of the cerebral cortex only (ctx-operated group) hampered the acquisition of limbic behaviors. There was no difference in cSI and LI between the ctx-operated and ablation groups, but there was a difference between ctx-operated and the unlesioned kindled group. There was also no difference in cSI between SC deep layer transection and ablation groups. Results of histologic analyses were similar for acute and kindled audiogenic seizure groups.
Conclusions: SC deep layers are involved in the expression of acute and kindled audiogenic seizure, and the cerebral cortex is essential for audiogenic kindling development.
Audiogenic seizures are induced by high-intensity sound stimulation (110–120 dB) in genetically susceptible rodents, and are widely used as a model of generalized tonic–clonic seizures (Kesner, 1966; Faingold, 1988; Jobe et al., 1991). The Wistar audiogenic rat (WAR, Doretto et al., 2003) is an inbred rodent strain susceptible to audiogenic seizures, which Ross and Coleman (2000) has reviewed, together with genetically epilepsy-prone rats (GEPRs), as a valid animal model of epilepsy.
The relationship between SNPr and SC has been widely studied in nongenetic models of experimental epilepsy such as electroshock (Iadarola & Gale, 1982; Gale, 1988) and intracerebral or systemic (Maggio & Gale, 1989) bicuculline injections. Lesions of deep layers of SC, within an area called the dorsal midbrain anticonvulsant zone (DMAZ) by Redgrave et al. (1992) decreased the severity of electroshock seizures. Bilateral microinjection of muscimol [γ-aminobutyric acid (GABA)A receptor agonist] in SNPr blocks or attenuates motor seizures in several experimental models of epilepsy, probably by disinhibition of the SC (Gale, 1985; Dean & Gale, 1989). The complete bilateral SC ablation abolished the anticonvulsant effect of muscimol infused bilaterally into SNPr in electroshock-induced seizures, a result which depends on activity of the nigro-tectal pathway. However, hampering other nigral efferents, such as the tegmental, thalamic, and striatal pathways, do not contribute to that effect (Garant & Gale, 1987). Furthermore, although the systemic application of phenobarbital is anticonvulsant in WARs, the microinjection of phenobarbital and muscimol in SNPr does not have anticonvulsant effects in audiogenic seizures (Rossetti et al., 2006).
Lesions of deep layers of SC in DBA/2 mice attenuated seizure activity (Willott & Lu, 1980), and microinjections of a N-methyl-d-aspartate (NMDA) antagonist into SC deep layers blocked reversibly audiogenic seizures of GEPR-3s (Raisinghani & Faingold, 2003). SC deep layers are important for the generation of wild running in GEPR-9s (Faingold & Randall, 1999), whereas microinjection of bicuculline in the SC of GEPRs induces audiogenic-like seizures (Merrill et al., 2003). Further evidence that SC is involved in the propagation of seizure activity was provided in GEPR-9s by in situ hybridization for c-fos mRNA (Ribak et al., 1994). More recently, Fuentes-Santamaría et al. (2007) showed morphologic and neurochemical abnormalities in SC deep layers in genetically epilepsy-prone hamsters (GPG/Vall). In contrast, in GEPRs Eells et al. (2004) and in Wistar rats from Strasbourg Simler et al. (1994) found no involvement of SC in acute and kindled audiogenic seizures.
Because the role of SC in several experimental models of epilepsy is still controversial, the present study was undertaken to examine the involvement of SC deep layers in acute and kindling audiogenic seizures in WARs. Briefly, we evaluated the long-term effects of transections disconnecting the deep layers of SC from the underlying tissues and the ablation of the SC areas corresponding to those above the transection in acute and kindled audiogenic seizures.
Male and female rats susceptible to audiogenic seizures taken from the WAR strain bred at the animal facilities of the Physiology Department of the Ribeirão Preto School of Medicine, University of São Paulo (Garcia-Cairasco et al., 1993; Doretto et al., 2003) and nonepileptic male and female rats (resistant) taken from the main breeding stock of the University of São Paulo were used. Animals weighing 250–350 g were housed five per cage, kept at 24°C, with access to food and water ad lib and with an artificial light–dark cycle of 12 h–12 h (lights on at 07:00 a.m., lights off at 07:00 p.m.). All animals were submitted to three screening tests at 70, 74, and 78 days of age. Only WARs displaying at least the tonic component of seizures; controls displayed no behaviors related to seizures (respectively cSI ≥ 4 and cSI = 0.00; see index below) in all of the three tests. Screening tests were conducted always between 5:00 and 7:00 p.m.. Efforts were made to avoid any unnecessary distress to the animals, in accordance with the Brazilian Society for Neuroscience and Behavior Guidelines for animal experimentation.
Sound stimulation and analysis of behavior
The testing apparatus consisted of a 32 × 36 cm transparent cylindrical cage, inside a larger, soundproof wooden box, both provided with doors and frontal glass windows. Animals were observed for 1 min before stimulation. Sound stimulus was provided by an electric door bell sound (120 dB), recorded on an audiotape, and delivered in the acoustic chamber through a loud speaker placed in the back wall of the larger box until tonic seizures appeared or for a maximum of 1 min.
Acute audiogenic seizures
Behavior was assessed by direct and systematic observation using a set of discrete behavioral categories (Ethnogram) most commonly used in the test situation. A categorized severity index (cSI) was described by Rossetti et al. (2006), modified from Garcia-Cairasco et al. (1996). Briefly, the index includes a graded linear scale that determines a range of seizure severity. Typically, resistant animals do not display any epileptic behavior after such stimulus, whereas WARs present running fits, jumping, and atonic falling (wild running), which is sometimes followed by tonic–clonic seizures. The cSI values for the most frequent behavioral sequences are shown in Table 1. Latencies, considered as the time interval between the beginning of the stimulus and the beginning of the first running fit, in seconds, were also measured in some experimental groups.
aCategories that are generally followed by hind-limb clonic convulsions (HCC).
One wild running
Wild running plus jumping
Two wild runnings plus jumping
Tonic convulsion (opisthotonus)
Tonic seizures plus generalized clonic convulsions
Head ventral flexion plus cSI 5
Forelimb hyperextension plus cSI 6a
Hind-limb hyperextension plus cSI 7a
Kindled audiogenic seizures
The severity of limbic seizures is described by a limbic index (LI), according to Racine (1972), where a score of 0 = immobility, 1 = facial automatism, 2 = head nodding, 3 = unilateral or bilateral forelimb clonus, 4 = bilateral forelimb clonus and rearing, and 5 = rearing, falling and generalized seizures. To study the role of SC in audiogenic kindling development, 20 sound stimuli were presented twice a day during 10 consecutive days at 9:00 a.m. and at 5:00 p.m., starting on the fourth day after the surgeries (see subsequent text).
Animals were anesthetized with 2,2,2-tribromoethanol (Sigma-Aldrich, Steinheim, Germany), 2.5% solution (w/v) at a dose of 1 ml/100 g of body weight, i.p. Sections of superficial and deep layers of SC were done by using an assembly built with a cannula and a stainless steel wire (Fig. 1A), according to stereotaxic coordinates (Paxinos & Watson, 1998), related to the bone, as follows: anterior-posterior (AP) = 0.6 mm behind the intersection between the longitudinal and the transverse (posterior) sutures; lateral (L) = 0.0 mm; ventral (V) = 3.0 and 2.0 mm, respectively, for deep and superficial SC layer sections. A bone fragment of about 1 cm2 was removed using a dental drill and maintained in saline solution until the end of the surgery. The surgical procedure was performed using a maneuver that allowed for the transections without causing bleeding, by displacing gently the venous sinus in a rostral direction. After reaching the desired stereotaxic coordinates, the knife was turned 90 degrees to the left side or to both sides to make, respectively, uni- and bilateral transections. The ablation surgery was conducted in a manner similar as that for the sections, but removing the collicular tissue and the cortex above it (ablation group) or the cortex only (ctx-ablation) by aspiration, through a vacuum pump, according to the following stereotaxic coordinates: AP = 2.2 mm behind the intersection between the longitudinal and the transverse (posterior) sutures; L = ±1.7 mm; and V = −5.4 mm related to the bone. After reaching the desired stereotaxic coordinates, the knife was turned 45 degrees to the left and then to the right side to perform a bilateral ablation. The knife used is shown in Fig. 1B.
In both kinds of surgery, the removed bone fragment was then returned to its place in the skull window, and the scalp was sutured. A few days later the bone fragment was completely adhered to the skull.
Transections and ablations of SC in kindled audiogenic seizure model
The experiments were conducted in WARs and Wistar resistant female rats, which were sound-stimulated twice a day, during 10 days, starting the set of stimuli at the 4th day after surgery in the following experimental groups:
1 WARs with bilateral SC deep layer transections (WARs, transection, n = 6).
2 Wistar resistant with bilateral SC deep layer transections (R, transection, n = 4).
1 WARs with opening the bone and dura (WARs, sham-operated, n = 6).
2 WARs with bilateral cortex removal (WARs, ctx-operated, n = 4).
3 WARs with bilateral cortex and SC ablation (WARs, ablation, n = 6).
4 Wistar resistant with opening the bone and dura (R, sham-operated, n = 5).
5 Wistar resistant with bilateral cortex removal (R, ctx-operated, n = 4).
6 Wistar resistant with bilateral cortex and SC ablation (R, ablation, n = 6).
Transections of SC in acute audiogenic seizure model
The experiments were conducted in male WARs, which were sound stimulated at 5, 10, and 15 days after surgery, in the following experimental groups:
1 Sham-operated: opening the bone and the dura mater only, but with no ablation or cuts in the nervous system (n = 7).
2 Inter-SC transection: transection between right and left SC (n = 8).
4 Unilateral transection reaching deep layers of SC (n = 8).
5 Bilateral transection reaching deep layers of SC (n = 7).
At the 16th day after surgery, animals from all groups were anesthetized and perfused transaortically with NaCl 0.9% solution and then with 10% formaldehyde solution. After at least 24 h of fixation, brains were cut into 20-μm sections, mounted on slides, and processed for Nissl staining. Lesion reconstructions were made using the diagrams of Paxinos and Watson (1998). The histology in all groups was performed 16 days after surgery, and it is known that this postlesion time-window may result in structural and functional plastic changes. Therefore, to control for eventual structural changes, one rat was submitted to SC deep layer transection and the histology was performed 24 h after the surgical procedure to show the actual acute lesion.
Data of the cSI, LI, and latencies of wild running represent means ± standard error (SE) of values obtained from groups of 4–8 animals, recorded in each experimental group, and were analyzed using the statistical program SIGMA STAT (Jandel Scientific, San Raphael, CA, U.S.A.). The test of normality showed that all data were nonparametric; therefore, the results were compared using Kruskal-Wallis analysis. Results were plotted with ORIGIN 5.0 (MicroCal Software, Northampton, MA, U.S.A.) and were considered statistically significant at a minimum confidence level of p < 0.05.
Effect of SC knife transections in the acute audiogenic seizure model
There was no statistically significant difference in cSI when the groups used as controls were compared (1, opening the bone and the dura mater; 2, transectioning between both SCs; and 3, unilateral transection reaching superficial layers of SC). In the experimental groups, animals which before surgery had high cSI presented only wild running behaviors (cSI = 2 or 3), 5 days after surgery. This behavioral pattern was maintained until the 15th day after surgery. The group with unilateral transections of SC deep layers presented abolishment of tonic–clonic seizures in all animals. The group with bilateral transections of SC deep layers presented abolishment of wild running and tonic–clonic seizures (cSI = 0) in three of seven animals at all evaluated times (5, 10, and 15 days after surgery) and presented cSI ≤ 2 in four of seven animals. There was no statistically significant difference between unilateral and bilateral SC deep layer transections, and both procedures differed significantly from control groups (Fig. 2A). When the animals presented seizures, the latencies of the first running fit episode were significantly higher in the group with bilateral transections of the deep layers of SC in all evaluated times (5th, 10th, and 15th day), compared to the group with unilateral SC deep layers. There was no difference among the other control groups (Fig. 2B).
After the wild running with jumping and atonic falling, usually WARs displaying cSI = 2 stopped and remained motionless until the end of sound stimulation (five of eight and four of seven in the unilateral and bilateral transections of the deep layers of SC groups, respectively). However, some animals displaying cSI = 2 (three of eight and three of seven in the unilateral and bilateral transections of the deep layers of SC groups, respectively) presented a particular kind of behavior. Once they started the wild running plus jumping and atonic falling, they continued circling without stopping until the end of the sound stimulation. We termed this behavior “persistent wild running.”
Effect of SC knife ablation and transection in the audiogenic seizure kindling model
WAR sham-operated group (opening the bone and the dura, with no transection) submitted to the protocol of audiogenic kindling development after surgery presented a tendency toward decreasing of cSI and a tendency toward increasing of LI as the number of stimuli increased (Fig. 3A–D). There was no statistically significant difference in cSI between sham-operated, transection, and ablation groups (Fig. 3A, C), and between ctx-operated, ablation, and sham-operated groups (Fig. 3B, D).
WAR sham-operated animals presented limbic behaviors during the development of the kindling protocol, reaching limbic class 4 (Figs. 3A and 3C, D). Transection group reached only limbic class 1 (Fig. 3C), whereas the ablation (Fig. 3A) and ctx-operated (Fig. 3B, D) groups did not present any limbic behaviors. There was no difference between ctx-operated and ablation for either cSI or LI, and there was also not significant statistical difference in cSI between transection and ablation groups (data not shown). Resistant animals did not display any wild running behavior or tonic–clonic seizures after either cerebral cortex ablations or transections (data not shown).
Because plasticity phenomena may occur 15 days after surgery, an acute transection lesion was done in a rat and the histologic examination was performed 24 h after surgery (Fig. 4A, B). Results of the histology analysis were similar for acute and kindling groups. Fig. 4C shows bilateral transections extending from 5.8 to 7.04 mm of the Paxinos and Watson (1998) interrupting almost completely the connections between SC deep layers and lower mesencephalic structures in those planes, touching slightly the limits of the periaqueductal gray matter. The most caudal portions of SC at the 7.3 mm plane were not reached. At the lateral plane, the transections were almost complete. Otherwise, the ablation group showed a very extensive lesion (cortex and hippocampus), as shown in a schematic drawing depicting the histology in the coronal, sagittal, and horizontal planes of one rat with SC deep layer ablation (Fig. 5). The corresponding cSI and LI values of this animal are also shown.
Unilateral or bilateral subcollicular (SC) deep layer transections blocked completely audiogenic seizures or significantly reduced cSI, whereas the latencies of the wild running, when still present, increased significantly only in the bilateral SC transection group. With similar results obtained with GEPR-3s and GEPR-9s subcolonies, Merrill et al. (2003) showed that only the lesion of deep layers of SC, in the DMAZ area, decreased the severity of audiogenic seizures. In addition, in recent studies, Rossetti et al. (2006) showed that the disinhibition of SC through inhibition of SNPr neither abolished nor decreased the audiogenic seizure severity of WARs.
Although microinjections of muscimol into the SNPr and picrotoxin into the SC did not block audiogenic seizures in Wistar rats from Strasbourg, bilateral injections of muscimol into the SC blocked the facilitation of audiogenic seizures induced by SC picrotoxin (Depaulis et al., 1990). Therefore, we can also infer that seizure control in genetically developed strains is different from those of nongenetic models of epilepsy (Gale & Iadarola, 1980; Garant et al., 1986; Maggio & Gale, 1989). In the latter models the deep layers of SC need to be inhibited or transectioned to display an anticonvulsant effect.
In GEPR-9s it has been shown that the decrease of audiogenic seizure severity during audiogenic kindling through lesions of SC or treatments with low-dose phenytoin increased the incidence of forebrain seizures. Animals that continued to display full tonic seizures did not exhibit forebrain convulsions, but did show posttonic clonus and forebrain seizure activity on EEG (Merrill et al., 2003, 2005).
There was no statistically significant difference in the cSI between SC transection and ablation groups during audiogenic seizure kindling stimuli. It is worth noting that in all experimental-lesioned groups (ablation, transection, and ctx-operated), the cSI started to decrease earlier than in the sham-operated group. Furthermore, the cSI in the operated groups presented lower variations compared to the sham-operated group.
Changes of LI showed that the cerebral cortex is essential for audiogenic kindling development, because only WARs from the sham-operated group (without ablation and ctx-operated) developed the audiogenic kindling phenomena. It is interesting to note that earlier studies have suggested a lack of involvement of cortex in the expression of acute audiogenic seizures (Kesner, 1966). However, the same author demonstrates a mild effect, on audiogenic seizures, of hippocampal lesions and a greater effect of caudate nucleus lesions. It is important to recall that in our histologic analyses that the so-called cortical ablation included sometimes hippocampus, which although not involved in acute audiogenic seizures, is strongly recruited during audiogenic kindling (Romcy-Pereira & Garcia-Cairasco, 2003). Similar data have been described in GEPRs in which repeated sound stimulation evoked neurochemical alterations (Lasley, 1991) and new limbic seizure components (Naritoku et al., 1992), which were abolished by amygdala microinjections of GABA agonists and NMDA antagonists (Feng et al., 2001). Furthermore, Raisinghani and Faingold (2005) demonstrated that the perirhinal cortex participates importantly in the neuronal network for audiogenic seizure as a result of audiogenic seizure kindling, and also demonstrated a previously unknown involvement of the perirhinal cortex in generalized onset seizures in GEPRs-3. Finally, extensive connections between the retrosplenial cortex and the SC are described as extremely important for sensorimotor integration, being the origin of SC-mediated motor and physiologic responses involved in emotional behavior or simply the link between the limbic system and the SC (Garcia del Caño et al., 2000). Because the retrosplenial cortex has been a possible target reached in the current extensive cortical ablations, we cannot rule out that its lesion can explain at least part of the effect of cortical ablation in audiogenic kindling.
The participation of SC in seizure expression is strongly related to the efferent pathway that links brainstem structures to spinal cord, throughout activation of motor neurons involved in the behavioral expression of seizures. Particularly, in audiogenic seizures it has been shown in GEPR-9s that the propagation of seizure activity from the IC involves activation of motor neurons in the nucleus reticularis pontis oralis (RPO, Browning, 1986) and that SC is important for the spreading of seizures activity between the IC and RPO. It is interesting to know also if the SC-reticular-spinal cord pathways in genetically developed audiogenic strains such as WARs or GEPRs are similar to those described, for example, for electroshock-induced seizures (Shehab et al., 2005, 2007).
The “persistent wild running,” is a striking change in the behavioral pattern observed in some animals after unilateral or bilateral SC deep layer transection from the audiogenic seizure groups. This finding is suggestive that the mechanisms responsible for stopping seizures were disrupted by SC subcollicular transections. Other explanations could be related to the disruption of normal activity of efferent projections from dorsal midbrain, which has been associated with defensive behaviors (Ribeiro et al., 2005; Castellan-Baldan et al., 2006). It is intriguing to note that Shehab et al. (2007) show a powerful anticonvulsant action often associated with locomotor activation. In their study, locomotor activation was characterized by compulsive locomotion, shuffling, circling, or rolling.
In additiion, data from our group show that the SNPr–SC circuitry might be involved with the end of seizures, because during hind-limb clonic seizures in WARs there is an increase in the main electroencephalographic frequency oscillation recorded in these nuclei (Rossetti et al., 2006). Probably, the disruption of SNPr–SC circuitry induced an alteration of a subcortical loop formed by basal ganglia, SC, and thalamus. In these loops, SNPr receives neuronal projections from the striatum (STR), and then projects to multisensory deep layers of SC, and from there to intralaminar thalamic nuclei and finally back to the STR (McHaffie et al., 2005). The capacity of these circuits to produce anticonvulsant effect, through the inhibition of SNPr, excitation of STR, or disinhibition of SC appears in several studies with animal models of epilepsy (Cavalheiro & Turski, 1986; Cavalheiro et al., 1987; Turski et al., 1987; Dean & Gale, 1989; Turski et al., 1989, 1990, 1991; Dean & Redgrave, 1992; Redgrave et al., 1992). Several lines of evidence have suggested that the SC may be an important output relay of an endogenous anticonvulsant circuit, especially in its connection with the SNPr (Nail-Boucherie et al., 2002). However, in audiogenic strains, our current results and data from the literature have suggested that this circuitry can be related to proconvulsant rather than to anticonvulsant mechanisms (Merrill et al., 2003; Rossetti et al., 2006).
Efferent projections from the dorsal midbrain have been associated with several functions, including defensive movements, which range from slight cringing to violent escape (Dean et al., 1988; Mitchell et al., 1988). It is possible that what these authors describe as a “violent escape reaction” is the same we described as “persistent wild running.” Additional studies in rats, however, have shown that the SC deep layers participate in flight-orientation responses (Sahibzada et al., 1986; Dean et al., 1989) and flight behaviors (Eichenberg et al., 2002), which are very similar to audiogenic seizure responses. It was not possible to find any relationship between the presence of this kind of behavior and the lesion extension seen in the histologic analysis.
In conclusion, our results showed that SC plays an important role in both acute audiogenic seizures expression and audiogenic kindling development. In addition, during audiogenic kindling development, in addition to SC, the cerebral cortex above the SC is also essential.
Maria Carolina Doretto was supported by CAPES, CNPq, and FAPEMIG. Franco Rossetti was supported by FAPESP fellowships (03/12420-2 and 03/01134-9). Norberto Garcia-Cairasco is supported by FAPESP (03/11381-3, 03/00873-2), FAPESP-Cinapce (2005/56447-7), CNPQ, PRONEX, CAPES-PROAP-PROEX, and FAEPA grants. Norberto Garcia-Cairasco holds a CNPq-Research fellowship. Special thanks to Marcelo Cairrão Araujo Rodrigues for the critical reading of the manuscript; and to Leonardo Fidelis Filho and Eduardo Gomes for the care of the animals at the Vivarium of the Physiology Department of the Ribeirão Preto School of Medicine, University of São Paulo.
We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.
In addition to that we have no conflict of interest to disclose.