14.1 Introduction

Learning and memory refers to the acquisition, consolidation, and retention of information for future use (retrieval). It enables an organism to “predict” future events, on the basis of this information and the current set of external and internal conditions, to adjust its behavior accordingly (Abel & Lattal, 2001). When this information is vital—for example, the recall of the location of a food source or mating partner, or the avoidance of a predator—it is obvious that the biological mechanisms of learning and memory are of considerable adaptive value for all species. Therefore, it is not surprising that several forms of information preservation can be found in animals (memory storage in the nervous and immune systems) and even plants (Rensing, Koch, & Becker, 2009).¹ The question, though, is whether these processes of information storage follow similar general rules, based on similar cellular and physiological processes. Such a general rule was postulated by Donald Hebb (1949, p. 62): “When an axon of cell A is near enough to excite a cell B and repeatedly or persistently takes part in firing it, some growth process or metabolic change takes place in one or both cells such that A´s efficiency, as one of the cells firing B, is increased.” It would be interesting to know which of these sorts of mechanisms share common ancestry and which represent parallel paths of evolution.

From a human perspective, it is interesting to ask what differences and similarities exist between learning mechanisms and memory systems of different organisms and how they relate to the human memory system, which is so important for our personality and individuality. This chapter gives a short overview of the neurobiology of learning and memory in different animal species, including both vertebrates and invertebrates. I put a special emphasis on the molecular basis of learning and memory across different time scales. However, memory is a complex phenomenon, involving multiple functions and brain systems, classified according to the way the information is stored and retrieved (see Figure 14.1; Henke, 2010; Kandel, Schwartz, & Jessell, 2000).

As depicted in Figure 14.1, human memory can be divided into short‐term (e.g. sensory and working memory) and long‐term memory, with long‐term subdivided further into explicit (declarative) and implicit (nondeclarative, including procedural) memory. Declarative memory can have episodic or semantic (remembering and knowing) characteristics. The existence of episodic memory in nonhuman animals has been disputed for a while, but now seems to be proven at least for some vertebrate species (Salwiczek, Watanabe, & Clayton, 2010). To the best of our knowledge—perhaps due to the experimental challenge of testing declarative memory in insects and snails—nothing is known about the existence of these memory forms in invertebrates. Hence, we here focus on nondeclarative memory.

14.2 General Aspects of Learning and Memory

Nondeclarative or implicit learning is broadly categorized into nonassociative and associative forms of information acquisition. In addition, imprinting, priming, and habit formation are observed in several species and are considered special forms of information storage (Horn, 2004; Yin & Knowlton, 2006). Nonassociative learning comprises habituation and sensitization, while associative learning is further divided into classical (Pavlovian) and instrumental conditioning (Kandel et al., 2000).

Habituation is the reduction in response magnitude (most often tested as a simple reflex, such as withdrawal or startle) due to repeated presentation of a stimulus. Sensitization refers to the enhancement or potentiation of a habituated response following a biologically relevant, strong stimulus. Habituation and sensitization are found in both vertebrates and invertebrates (Davis, 1984; Kandel, 2004).

In classical associative delay conditioning, the temporal pairing of a biologically relevant unconditioned stimulus (US) with a neutral conditioned stimulus (CS) leads to prediction that the CS will be followed—with a certain probability—by the US. Control experiments include the nonpaired condition, where the US and the CS are presented in an unrelated way. Intermediate between paired and unpaired training conditions are the so called trace conditioning procedures where the US follows the CS after a certain time gap. Here, the predictive value of the CS is determined by the duration of this temporal lag. In general, the predictive value of the CS changes the individual’s behavior to approach predicted rewards and avoid predicted aversive situations. Repeated presentation of the CS in the absence of the US after conditioning reduces the ability of the CS to elicit a conditioned response (extinction). On the other hand, repeated presentation of the prospective CS before conditioning reduces the degree of CS‐US association (latent inhibition). It is not possible in this short chapter to consider in depth all the different temporal aspects of learning (acquisition, consolidation, retrieval) and memory (short‐term, intermediate‐term, long‐term, permanent) when comparing the mechanisms between different species. Therefore, we focus on the mechanisms of acquisition and long‐term memory in classical conditioning.

A key cellular mechanism involved in associative learning and memory is long‐term potentiation (LTP). First described in 1973 by Bliss and Lomo in the rabbit hippocampus (1973), LTP is now well established as a process by which certain changes in synaptic activity can induce a long lasting and specific increase of the strength of this synapse in vertebrates (Lynch, 2004; Malenka, 2003; Morris, 2003) as well as in invertebrates (Glanzman, 2008; Menzel & Manz, 2005; Roberts & Glanzman, 2003).

14.3 Learning and Memory in Invertebrates

Neurobiological research in invertebrates has the advantage of dealing with relatively “simple” nervous systems, even allowing experimenters to target specific identified neurons. Perhaps the most eminent early studies into the molecular basis of learning and memory were started in the 1960s and 1970s by Eric Kandel and colleagues. He used the marine snails Aplysia depilans and A. californica for experiments on nonassociative and associative learning (Kandel & Pittenger, 1999; Kupfermann & Kandel, 1969). In this model, the gill‐withdrawal reflex is used as a dependent variable which is reduced in magnitude (habituation) after repeated stimulation with a mild water‐jet shot onto the skin of the snail. Conversely, the reflex is enhanced after noxious stimulation of the animal’s tail (sensitization) or after presentation of an otherwise neutral CS (e.g., mechanical stimulation of the mantle shelf) that has been paired with a noxious tail shock. Habituation and sensitization of the gill‐withdrawal reflex in Aplysia have been shown to involve, respectively, reduced or enhanced calcium‐dependent glutamate release from the presynaptic terminal of sensory neurons, which contact motor neurons that activate the gill muscles and mediate the reflex. Enhanced release is due to heterosynaptic facilitation by serotonergic interneurons. The stimulatory effects of serotonin (5‐HT, 5‐hydroxytryptamine) on the presynaptic terminal of the sensory neurons are due to activation of different types of calcium channels. This triggers a cascade of intracellular events involving cAMP‐dependent Protein Kinase A (PKA) and the stimulation of Protein Kinase C (PKC) by diacylglycerol. These processes are probably convergent to habituation and sensitization in vertebrates (Weber, Schnitzler, & Schmid, 2002).

However, postsynaptic signalling events triggered by a rise in intracellular calcium are also important for learning in Aplysia. For example, the induction of retrograde messengers such as the gaseous messenger molecule, nitric oxide (NO), (Michel, Green, Eskin, & Lyons, 2011) and the up‐regulation of the glutamate receptor, AMPA (α‐amino‐3‐hydroxy‐5‐methyl‐4‐isoxazolepropionic acid), in the postsynaptic motor neuron (Glanzman, 2008) are found in Aplysia. Structural changes of the synapses, the basis of long‐term memory in Aplysia, also involve cAMP‐response Element Binding Protein (CREB)‐mediated gene expression (Bailey & Kandel, 2008; Kandel, 2004).

Others have considered honeybees (Apis mellifera) as useful animals for the study of learning and memory (Bitterman, 1976; Lindauer, 1970; Menzel, 1979). Here, the appetitive proboscis‐extension reflex to olfactory food stimuli is often taken as the behavioral measure of these cognitive functions. The mushroom bodies, which comprise large parts of the bees´ brains, integrate various sensory pathways and are essential for learning and memory. In addition to the mushroom bodies, the antennal lobes have also been found to be involved in memory storage in bees (Giurfa & Sandoz, 2012; Hammer & Menzel, 1998; Menzel, 2012). Learning and memory formation in bees is also divided into different temporal phases and the molecular determinants of these different cellular processes are distinguished (Müller, 2002) resembling those outlined for humans in Figure 14.1. As in other animals, the initial cellular mechanisms involved in learning and memory also include a synapse‐specific rise in calcium levels, followed by activation of retrograde messengers such as NO (Müller, 1994, 1996), and finally, activation of several different protein kinases (Giurfa, 2007; Menzel, 2001). For long‐term memory it appears that, in honeybees, CREB (or rather its bee orthologue, AmCREB) plays the essential role (Eisenhardt et al., 2003; Müller, 2000).

In addition, research on the fruitfly (Drosophila melanogaster) has pioneered the neurogenetic basis of learning and memory in invertebrates, especially with respect to synaptic physiology (Davis, 2011; Dubnau & Tully, 1998; Heisenberg, 2003; Heisenberg, Borst, Wagner, & Byers, 1985). As in other memory systems, the initial signal relevant for the selective strengthening of specific synapses is an increased calcium influx into Kenyon cells of the fly’s mushroom bodies. The sequel of this event includes the activation of NO‐synthase, adenylate cyclase, increased intracellular cAMP levels and subsequent activation of protein kinases (Gerber, Tanimoto, & Heisenberg, 2004; Heisenberg, 2003; Müller, 1997). Here, as in other species, long‐term memory formation appears to require CREB‐dependent gene transcription and protein synthesis in a subset of neurons (Chen et al., 2012).

14.4 Learning and Memory in Vertebrates

Most vertebrate learning studies are done in mammals, especially rodents (rats and mice) and humans. Therefore, we shall focus on mammalian memory systems here, despite the fact that there are fascinating forms of learning and memory in nonmammalian models, including imprinting and song learning in birds, as well as the formation of declarative memory in various species.

Several brain systems that are relatively conserved across mammals appear to be adapted for learning and memory (probably through anatomical and neurochemical characteristics) including amygdala, hippocampus, cerebellum, basal ganglia, and cerebral cortex (Henke, 2010). This list is by no means complete. For example, consider the formation of long‐term pain memories in the mammalian dorsal horn of the spinal cord, associated with hyperalgesia. However, of these regions, the hippocampus is the most famous. Since Scoville and Milner (1957) reported severe anterograde amnesia after bilateral temporal lobe resection in their patient, H.M. (Henry Gustav Molaison, 1926–2008), and especially since the discovery of LTP (Bliss & Lomo, 1973), the hippocampal formation has been the paradigmatic brain structure in which to study learning and memory processes in mammals (van Strien, Cappaert, & Witter, 2009).

In these brain structures, several functional and morphological modifications occur in the course of learning, mainly triggered by an increase in postsynaptic entry of calcium into the cell through NMDA (N‐methyl‐D‐aspartate) glutamate receptors and voltage‐gated calcium channels (Wang, Hu, & Tsien, 2006). The NMDA receptor is an ion channel comprised of four subunits that is mainly permeable for sodium and calcium. This channel has the very interesting property of being both ligand‐ and voltage‐gated. At membrane potentials of up to −35 mV the channel pore is blocked by a magnesium ion. However, once the membrane is depolarized > −35 mV, Mg²⁺ is expelled from the channel, so that sodium and calcium can enter the neuron. (Due to the special properties of being ligand‐ and voltage‐gated the NMDA receptor has been termed a “coincidence detector,” because the opening properties depend on the temporally coincident input.) This rise in the intracellular calcium concentration triggers short‐ and long‐term changes in second‐messenger systems. One might ask how a general increase in the postsynaptic Ca²⁺ concentration can lead to rather specific intracellular effects, ultimately strengthening a particular synaptic input. This is mainly due to the restriction of diffusion of the ion by mechanical barriers in the dendritic spines (nano‐ and microdomains), by binding to different Ca²⁺‐binding proteins and through differences in the kinetics of Ca²⁺ channels (Burgoyne, 2007).

Short‐term plasticity includes retrograde signaling, mostly through NO, which once synthesized from l‐arginine, diffuses back to the presynaptic terminal to activate guanylate cyclase, increase the

Related posts:

The Brain Evolved to Guide Action Social Coordination: From Ants to Apes Neurotransmission—Evolving Systems Social Learning, Intelligence, and Brain Evolution Neural Development in Invertebrates Nervous System Architecture in Vertebrates

Stay updated, free articles. Join our Telegram channel

Join