- Top of page
- Materials and methods
Chick imprinting behavior is a good model for the study of learning and memory. Imprinting object is recognized and processed in the visual wulst, and the memory is stored in the intermediate medial mesopallium in the dorsal pallium of the telencephalon. We identified chicken cholecystokinin (CCK)-expressing cells localized in these area. The number of CCK mRNA-positive cells increased in chicks underwent imprinting training, and these cells expressed nuclear Fos immunoreactivity at high frequency in these regions. Most of these CCK-positive cells were glutamatergic and negative for parvalbumin immunoreactivity. Semi-quantitative PCR analysis revealed that the CCK mRNA levels were significantly increased in the trained chicks compared with untrained chicks. In contrast, the increase in CCK- and c-Fos-double-positive cells associated with the training was not observed after closure of the critical period. These results indicate that CCK cells in the dorsal pallium are activated acutely by visual training that can elicit imprinting. In addition, the CCK receptor antagonist significantly suppressed the acquisition of memory. These results suggest that the activation of CCK cells in the visual wulst as well as in the intermediate medial mesopallium by visual stimuli is indispensable for the acquisition of visual imprinting.
In visual imprinting, day-old chicks rapidly form a memory of the moving object to which they were first exposed and show a following response (Lorenz 1937). The importance of the intermediate medial mesopallium [IMM, all nomenclatures are in accordance with Reiner et al. (2004)], which is a subregion of the dorsal pallium region, for the acquisition and/or retention of memory in visual imprinting has been suggested by many studies (Horn et al. 1979; Kohsaka et al. 1979; McCabe et al. 1981, 1982). In addition, we have reported that the visual wulst, which is located dorso-rostrally to the IMM, is associated with the acquisition of visual memory in imprinting and exhibits plastic changes (Maekawa et al. 2006). The visual wulst is the highest center of the thalamofugal pathway in the telencephalon, which is one of two pathways for visual information processing in birds (Karten and Hodos 1970; Karten et al. 1973; Benowitz and Karten 1976).
What are the cellular and molecular mechanisms underlying this plasticity? It is known that learning during the early post-natal periods in Aves induces several biochemical and structural changes in the IMM, including an increase in the size of the post-synaptic densities (Bradley et al. 1981; Horn et al. 1985), learning-related increases in the phosphorylation of the myristoylated alanine-rich C-kinase substrate, an increase in the activation of calmodulin-dependent protein kinase II (Sheu et al. 1993; Zhao et al. 1999), an increase in the number of vesicles near the active release zones in synapses (Ruskov et al. 1995), and an increase in the glutamate, GABA, or taurine release (Tsukada et al. 1999; McCabe et al. 2001; Meredith et al. 2004). All these events may facilitate synaptic transmission, and indeed, the proportion of neurons in the IMM that responded to the imprinted stimulus was increased in imprinted chicks (Nicol et al. 1995).
In addition to glutamate and GABA, several neuropeptides are found in the mammalian neocortex, some of which may play important roles in the learning process. Although their functions have not yet been fully elucidated, these peptides can modulate the activity of neurons both post-synaptically through peptidergic action on receptors and pre-synaptically via the release of glutamate or GABA.
In the present study, we focus on the roles of these peptides in the imprinting behavior in order to reveal a novel regulator of memory acquisition. We attempted to identify the types of neurons in the dorsal pallium that are excited by visual stimuli during the training for visual imprinting, using a combination of c-Fos immunohistochemistry and the detection of various transmitters by immunohistochemistry and/or mRNA in situ hybridization. As a result, we found that more than half of the cholecystokinin (CCK) neurons in the dorsal pallium were also immunopositive for c-Fos after imprinting training. Therefore, we performed detailed examination and characterization of these CCK neurons in order to clarify their roles in inducing visual imprinting.
- Top of page
- Materials and methods
Visual imprinting, which is a form of filial imprinting in which chicks learn the ensemble of characteristics presented by the mother hen, is an instinctive behavior of ground-nesting fowl that was first analyzed by Konrad Lorenz in the 1930s (Lorenz 1937). Previously, we have shown that the visual wulst, which is located in the dorsal part of the pallium, is critical for the acquisition but not the recall of memory in visual imprinting behavior and exhibits plastic changes after imprinting training (Maekawa et al. 2006). From these results, we postulated that there should be a cellular correlation that supports this plasticity in the dorsal pallium. In the current study, we found that CCK expression and c-Fos immunoreactivity colocalized in the pallium after imprinting training. We confirmed that the CCK mRNA level was increased after imprinting training using both quantitative real-time PCR methods and counting the number of cells that expressed CCK mRNA. These results indicate that many CCK neurons in the dorsal pallium become active following visual imprinting training. Furthermore, we demonstrate that these changes are specifically related to the acquisition of imprinting, as they were not observed in chicks that were trained after closure of the critical period. The indispensability of CCK signaling for the acquisition of visual imprinting was further confirmed using devazepide, which is a potent antagonist of chCCK-R (Nilsson et al. 2003).
We also characterized the CCK neurons in the dorsal pallium. For this purpose, we used GAD65 as a GABAergic marker, VGLUT as a generic glutamatergic marker (Fremeau et al. 2001; Cheng et al. 2004), and parvalbumin, which is a neuropeptide that is known to be expressed in interneurons. Only a small subset of the CCK neurons co-expressed GAD65 mRNA, which indicated that they are inhibitory interneurons. However, most of the CCK neurons were glutamatergic, as judged from VGLUT2 expression, and were devoid of parvalbumin immunoreactivity. The glutamatergic property of most of the CCK neurons was unexpected, as CCK is usually considered to be one of the inhibitory interneuron markers in mammals (Somogyi et al. 1984; Markram et al. 2004; Somogyi and Klausberger 2005). However, other studies have demonstrated that the expression of CCK mRNA is more widespread than was previously assumed and occurs in both the interneurons and the pyramidal cells of the rodent cortex, the latter generally being thought to be glutamatergic (Burgunder and Young 1990; Schiffmann and Vanderhaeghen 1991; Senatorov et al. 1995; Gallopin et al. 2006). Therefore, we assumed that glutamatergic CCK neurons were common in the pallium, and that there may be a species difference in the frequency of occurrence. The precise distribution and functional role of inhibitory CCK neurons in the chick pallium must be determined in future studies.
Although chCCK-R was expressed in the region in which the CCK cells were observed, the precise physiological consequence of CCK release at the cellular level in the chick dorsal pallium is not known at present. In mammals, CCK increases GABA release in the cortex and hippocampus, leading to an increased tonic inhibition of pyramidal neurons (Sheehan and de Belleroche 1983; Perez de la Mora et al. 1993; Miller et al. 1997). Alternatively, activation of the CCK-B receptor, which is a major receptor expressed in the mammalian brain (Gaudreau et al. 1983; Honda et al. 1993), induces phosphatidylinositol hydrolysis and intracellular Ca2+ mobilization (Kopin et al. 1992; Shinohara and Kawasaki 1994; Shigeri et al. 1996). As chCCK-R has a high affinity for sulfated CCK-8, CCK-4, and sulfated gastrin-17 (Nilsson et al. 2003), it may have a function similar to that of the mammalian CCK-B receptor. Therefore, it is possible that these CCK neurons have an excitatory effect on post-synaptic neurons that express chCCK-R. In addition, CCK can facilitate glutamate release in rat hippocampal neurons (Migaud et al. 1994; Breukel et al. 1997). These excitatory effects of CCK are consistent with the fact that CCK increases the excitability of pyramidal neurons in the cortex as well as in the hippocampus in mammals (Dodd and Kelly 1981; Jaffe et al. 1987; Boden and Hill 1988; Bohme et al. 1988; Shinohara and Kawasaki 1997; Gallopin et al. 2006). Further physiological characterization at the single cell level is necessary to elucidate the precise role of CCK neurons in the chick dorsal pallium.
A substantial body of evidence suggest that CCK is important in the regulation of anxiety, pain, food intake, memory, and other activities in the central nervous system (Crawley and Corwin 1994). These various biological effects are mediated by the widespread distribution of CCK-R in the brain, although the functional relationships between these various biological effects are not clear at present. However, it is conceivable that CCK, which has a feeding-suppressing effect, facilitates learning and memory, as do other gut-brain peptides, such as bombesin (Ohki-Hamazaki et al. 2005).
To our knowledge, this is the first report to show clearly the involvement of CCK in visual learning. In the murine system, visual learning is difficult to evaluate because of the poor visual discriminating abilities of these animals. However, our results are reminiscent of other studies in rodents. A selective CCK-B agonist improved cognitive performances evaluated by habituation to a novel environment or two-trial recognition memory task in rats (Gerhardt et al. 1994; Lena et al. 1999; Sebret et al. 1999; Taghzouti et al. 1999). In the latter case, involvement of the hippocampus was further shown by the local injection of an agonist and antagonist and by an increase in the extracellular levels of CCK-like immunoreactivity in the hippocampus (Sebret et al. 1999). In line with these reports, the enhancing effects of CCK on certain types of learning and memory are evident.
In conclusion, CCK neurons are abundant in the chick dorsal pallium and are excited by a visual stimulus that evokes visual imprinting. Inhibition of chCCK-R impairs the acquisition of memory, which is indispensable for visual imprinting. These results indicate for the first time that the activation of chCCK-R is critical for memory formation in imprinting, and CCK may have an important role in shaping neural networks for imprinting.