Invertebrates as models of learning and memory: investigating neural and molecular mechanisms

What do the experiments that revealed the cellular components of the brain (Ramón y Cajal, 1894; Garcia-Lopez et al., 2010), described the mechanisms of nerve impulse transmission (Castellucci and Kandel, 1974; Hodgkin and Huxley, 1952; Katz, 1949, 2016), led to the identification of the neurotransmitter gamma-aminobutyric acid (Florey, 1991), and led to the characterization of the molecular basis of learning and memory have in common?

They have all marked the history of neuroscience, have earned numerous Nobel Prizes, and, not least, were all conducted in invertebrate organisms.

Although invertebrates possess small nervous systems, consisting of a limited number of neurons or ganglia, they are not limited in their ability to produce sophisticated and complex behaviors or even high-order forms of learning (see Glossary) (Preuss, 1995). Indeed, many invertebrates possess large neurons which facilitated microelectrode recordings and allowed the characterization of the neural and molecular basis of learning and memory in a comparative context (Kandel and Kupfermann, 1970).

The evolutionary process that prompted the diversity among species also promoted the conservation of numerous key physiological processes that are well preserved across taxa (Pembroke et al., 2021). By virtue of this, although invertebrates possess small brains and are phylogenetically distant from mammals, they have been and still are of fundamental importance in understanding basic neuroscience and in accelerating the pace at which mammalian studies can be translated to humans (Fig. 1) (Rivi et al., 2020). While maintaining the simple organization of the invertebrate nervous system, the behavioral repertoires and cognitive abilities of mollusks and arthropods have been shown to be highly comparable to those of mammals (Benjamin et al., 1985; Carew et al., 1981; Crow and Alkon, 1978, 1980; Lederhendler and Alkon, 1987; Strausfeld, 2012). This unique combination of simple nervous systems and complex behaviors provided a major contribution to the characterization of the conserved mechanisms by which memory is formed and stored. Thus, the results obtained in invertebrates have been translated first to mammalian models and then to humans.

Note that this Commentary is not intended to be an exhaustive review of the invertebrate model systems that have been used to study the behavioral, cellular and molecular mechanisms of learning and memory. Owing to space limitations, we have restricted our discussion to selected mollusks and insects (especially flies and bees) as well as crayfish and planaria, but other extremely innovative work on learning and memory mechanisms using models such as Caenorhabditis elegans and cephalopod mollusks is worth mentioning and has been recently reviewed by Rahmani and Chew (2021) and Schnell et al. (2021).

The ability to form memories has profound effects on an organism's life and survival (Nairne and Pandeirada, 2016). The knowledge of past experiences allows animals to plastically respond to present challenges and thus promote adaptation to ever-changing environments (Bisaz et al., 2014). Given the importance of memory, many attempts have been made to characterize the cellular and molecular processes involved in cognitive functions and to localize the physical trace of a memory, which is known as an engram (Dudai and Eisenberg, 2004).

The idea that memory is stored as lasting changes in the brain dates back at least to Plato and Aristotle's time (∼2400 years ago), but its scientific articulation emerged in 1894 when Cajal first proposed that memory is stored as an anatomical change in the strength of neuronal connections (Bailey et al., 2000). For the following 50 years, little evidence was gathered to support this idea, until Hebb in 1949 proposed a model that memory is established through changes in the number and strength of synaptic connections between neurons (Hebb, 1949). This theory paved the way for studies performed in the 1960s and early 1970s aimed at investigating how changes in behavior resulting from training procedures reflect changes in the activity of specific neurons.

Whereas the majority of the scientific community was focused on characterizing the memory-related processes in rodent models, brilliant and revolutionary researchers such as Tauc and Kandel made the radical decision to focus their research on an invertebrate model, the sea slug Aplysia californica. They believed – correctly – that this model would make the study of the fundamental and conserved neuronal and molecular events associated with memory and learning easier. In the chapter ‘Searching for an ideal system to study memory’ of his 2007 book, Kandel (2007) defines his approach as ‘reductionist’, aspiring to replicate Pavlov's conditioning procedures in A. californica. Faria (2020) explains how, by taking advantage of the direct link between synaptic plasticity and behavior, Kandel characterized the neuronal networks and molecular cascades of learning and memory. These aspirations were later translated into results that earned Kandel, Carlsson and Greengard the Nobel Prize in Physiology or Medicine for their discoveries on the conserved mechanisms of synaptic plasticity (Carew et al., 1971; Kandel et al., 2014).

Over the past decades, this model contributed to new evidence about learning and memory processes and their alterations in pathological conditions (Abrams, 2012; Hawkins, 2013). For example, the molecular mechanisms that mediate the attention-like process in Aplysia have proved to be highly conserved, thus helping in characterization of attentional processes in mammals (Hawkins, 2013). Aplysia also provides a unique model to examine the effects of age on learning, memory and arousal. In fact, it has been demonstrated in Aplysia that aging: (1) impairs the long-term retention of habituation (see Glossary), (2) prevents the acquisition of sensitization in the siphon withdrawal reflex (see Glossary) and (3) reduces arousal, reflecting the age-dependent alterations of behavioral plasticity reported in vertebrates, including Homo sapiens (Bailey et al., 1983). Furthermore, as the molecular mechanisms underlying the effects of age on behavior share common features across phyla, the short lifespan of Aplysia and the simplicity of its central nervous system (CNS), represent important advantages for neuro-aging studies (Kron et al., 2020).

Inspired and encouraged by the early results obtained using A. californica, multitudes of neuroscientists all around the world exploited the ‘simpler’ nervous systems of invertebrates to define the synaptic and integrative properties of neural circuits that control behaviors and they are conditioned (Nader, 2015). Subsequently, an increasing number of invertebrate models entered the learning and memory field.

In 1950, Lashley published ‘In Search of the Engram’, a famous document in which he summarized his theories on memory and the brain (Bruce, 2001; Dudai and Eisenberg, 2004). Since that time, many questions remain unsolved and neuroscientists are still searching for the engram, the site of memory storage.

Despite the unquestionable importance of synaptic plasticity for brain function, the exact role of cellular and connectivity modifications in learning processes remains obscure (Humeau and Choquet, 2019). Moreover, despite recent progress in developing techniques for identifying and manipulating engrams at the neuronal level, the neurobiological underpinnings of memory retrieval remain almost unexplored (Frankland et al., 2019).

Given these unresolved questions, molluscan unipolar neurons represent a unique platform to investigate where memory is stored. In fact, their primary neurites, which are the site where most of the synaptic interactions occur, can survive for long periods without their soma and still remain competent to synthesize new proteins (Lukowiak et al., 2003; Scheibenstock et al., 2002). This finding allowed researchers to distinguish between the sites (neurites or soma) in which memories are processed (Kandel, 1979; Lukowiak et al., 2003). Whereas intermediate term memory (ITM) is only dependent on new protein synthesis, long-term memory (LTM) depends on both altered gene activity and new protein synthesis (Sakakibara, 2008). Thus, by ablating the soma, it has been possible to demonstrate that memories are consolidated and stored in the neurons' soma (Lukowiak et al., 2000; Sangha et al., 2003; Scheibenstock et al., 2002). Experiments like these have been conducted extensively in another molluscan organism: the great pond snail Lymnaea stagnalis. The results obtained were extraordinary, making it a better model than its ‘cousin’ Aplysia for the search for the engram. The reasons for this are threefold. First, the neural circuits that mediate interesting and tractable behavior, such as feeding and aerial respiration, have been characterized (Benjamin, 1983, 2012; Lukowiak et al., 2006; McComb et al., 2003; Spencer et al., 1999; Straub et al., 2004; Syed et al., 1992a,b). Second, these behaviors can be classically and operantly conditioned (see Glossary) and, depending on the training procedure used, both ITMs and LTMs can be formed (Batabyal et al., 2021; Benatti et al., 2020, 2022; Kawai et al., 2004; Kemenes and Benjamin, 1994; Lukowiak et al., 1996; Rivi et al., 2020, 2021a,b, 2022b,c,d,e). Third, for aerial respiration, a single neuron, RPeD1 (right pedal dorsal 1), has proven to be a sufficient and necessary site for memory formation, consolidation, reconsolidation and extinction (Sangha et al., 2003; Scheibenstock et al., 2002). This is, so far, the only instance in both invertebrates and vertebrates where a single neuron meets both the sufficiency and necessity criteria for mediating different hierarchical aspects of memory. In fact, if the soma of RPeD1 is ablated and the primary neurite is left behind (allowing the organism to perform aerial respiration), learning occurs and ITMs are formed, but LTMs can no longer be demonstrated (Scheibenstock et al., 2002). However, if the soma of RPeD1 is ablated after LTM consolidation, memory is still present, suggesting that the soma of RPeD1 is not needed for the retention of LTMs (Scheibenstock et al., 2002). These results have laid the foundation for future genomic and proteomic studies aimed at elucidating the molecular events occurring in this neuron, giving a significant contribution to defining the engram. Experiments such as these cannot be carried out in mammal preparations, as the disruption of the neuronal soma usually causes the death of the entire cell.

Another important challenge for neuroscience is the characterization of the neural circuits responsible for animal learning and behavior. The electron microscopy techniques available today, by enabling high-quality and multi-scale neuronal imaging, significantly advanced the understanding of brain-wide connectivity (Li et al., 2020) and showed that the learned changes in behavior following conditioning are the result of changes in connection strength between neurons across multiple circuits (Abraham et al., 2019). Such studies aim to generate a complete ‘map’ of the chemical synapses between all neurons in order to study the effects of the perturbation of single cells on behavior and physiology (Josselyn and Frankland, 2018; Sehgal et al., 2018). However, the development of a comprehensive understanding of these circuits at a single neuronal level cannot be easily performed in mammals because of the complexity of their brains and behaviors (Eichler et al., 2017; Li et al., 2020).

On the other hand, the fruit fly Drosophila melanogaster has proved to be a leading candidate to study the comprehensive structure and function of the brain, and the mechanistic basis of learning, memory formation and complex behaviors (Scheffer and Meinertzhagen, 2021). This is, in part, due to the simplicity of its CNS (Raji and Potter, 2021). Furthermore, a century of work on fly genetics makes manipulation of its ‘small’ brain easier than in any other animal species (Sokolowski, 2001).

To date, scientists are on their way to compiling a connectome (i.e. a map of all neurons and their chemical synapses) of the mushroom body of D. melanogaster (Eichler et al., 2017). The mushroom body is a higher-order parallel fiber system that is essential for flies to form and retain associations between stimuli and reinforcement (Eichler et al., 2017). Its conserved neuronal architecture and important role in learning and memory (Devineni and Scaplen, 2022; Shinomiya et al., 2022; Simpson, 2009) allowed the reconstruction of a circuit map of the mushroom body. This map will guide future comparative studies aiming to provide a functional understanding of how flies learn, remember and forget. When, in 1850, Dujardin first described mushroom bodies of insects and compared them with the vertebrate cerebral cortex, he could hardly have imagined that nearly 150 years later it would be shown that these structures are indeed involved in mediating conserved cognitive functions or that 170 years later the study of connectomes would be established.

The contribution made by invertebrates extends beyond the mere characterization of the ‘where’, ‘how’ and ‘when’ of memory formation, consolidation and loss. The applications of these models, in fact, extend on several fronts. For example, aquatic mollusks with their open circulatory systems allow the use of membrane-permeant drugs and compounds that can be easily absorbed. This has helped to unravel the complexity of various signaling pathways and provide new insights into how drugs and molecules can modulate different neuronal functions and behaviors (Fodor et al., 2020a,b; Rivi et al., 2020). In the past decades, many studies have demonstrated the memory-enhancing or -impairing effects induced by the exposure of organisms to drugs and compounds (Gho and Ganetzky, 1992; Monleón et al., 2008; Nakai et al., 2022; Søvik et al., 2018). In this context, L. stagnalis has been recognized as a useful organism to examine the effects of drugs such as cannabinoids, fluoxetine, methamphetamine, cocaine and propranolol as well as bioactive compounds, such as flavonoids, on learning and memory (Batabyal and Lukowiak, 2021; Benatti et al., 2017; Carter et al., 2006; Fernell et al., 2016; Il-Han et al., 2010; Ito et al., 2014; Kagan et al., 2022; Kennedy et al., 2010; Rivi et al., 2021a, 2022a; Swinton et al., 2018, 2020, 2021).

Furthermore, L. stagnalis has been used to study the effects of amyloid-β (Aβ) peptides on the snails' ability to form LTMs (Ford et al., 2015, 2017). Aβ peptides, in fact, are implicated in memory loss, neuronal impairment and neurodegeneration in Alzheimer's disease (Harrington, 2012) and Aβ1–42 oligomers have been identified as toxic fragments that likely affect LTM through synaptic plasticity pathways (Ford et al., 2015, 2017). These findings added an important piece to the puzzle for the global understanding of neurophysiological processes underlying aging and memory decline (Fodor et al., 2021).

Another very important taxon for the study of memory and learning is insects. Over the past few years, several studies have described the numerous ways in which insects have contributed to answering outstanding questions related to complex behaviors. For example, the honeybee (Apis mellifera) has been critically important for characterization of circadian rhythms (Rubin et al., 2006); the worm Caenorhabditis elegans, the planarian Dugesia japonica and the tussock moth (Eloria noyesi) have been adopted for addiction research (Søvik and Barron, 2013); lobsters (Homarus americanus) and crickets (Gryllus bimaculatus) can be used for studying aggressive behaviors (Briones-Fourzán et al., 2015); and many insect species provide unique systems to investigate how early-life experience alters the brain and behavior (Westwick and Rittschof, 2021). Research on insects has revealed the existence of a variety of cognitive phenomena and higher forms of learning and memory (see Glossary) that were previously thought to be restricted to vertebrates and, sometimes, only to humans. We provide examples of these here.

Honeybees can rapidly master two abstract concepts simultaneously using spatial relationships (above/below and right/left) and then transfer their choices to unknown stimuli that offer the best match in terms of dual-concept availability (Avarguès-Weber et al., 2012).

Studies on bumblebee (Bombus terrestris) and honeybee color learning provided the first evidence of attentional processes in insects and demonstrated how these complex mechanisms can be modulated by experience (Dyer and Garcia, 2014; Gumbert, 2000). These animals, in fact, can be trained to distinguish between a rewarded and a non-rewarded color as well as between target and distractor stimuli (Menzel and Giurfa, 2001). A recent study in D. melanogaster demonstrated that, in the presence of competing percepts, attention can be switched from one attentional state to another one. This phenomenon – known as ‘attentional rivalry’ – seems to be evolutionarily conserved. As the slowing of rivalry rate is associated with heritable psychiatric disorders, such as bipolar disorder, perceptual rivalry in flies represents a powerful model for investigating the genetic and molecular influences on rivalry rate and may even shed light on human cognitive and behavioral dysfunction (Miller et al., 2012).

Furthermore, taking advantage of the tractability of D. melanogaster in both behavioral and microbiome studies (Wong et al., 2014), Silva et al. (2021) recently tested how the elimination of microorganisms affects the organism's behavior. The study demonstrated that microbiologically sterile (axenic) flies had a moderate reduction in memory performance. Moreover, axenic flies showed a tendency to sleep longer and had reduced sleep rebound after sleep deprivation (Silva et al., 2021).

As growing evidence suggests, neural circuits conserved between the D. melanogaster and mammalian brain control not only wakefulness and activity but also many aspects of interactions between organisms and their gut microbiome (Cryan et al., 2019; Ezenwa et al., 2012). Thus, this model organism offers a great opportunity to elucidate the mechanisms underlying microbiome-dependent traits, opening a new avenue for translational studies.

Shifting from flies to crickets (Gryllus bimaculatus), Matsumoto and Mizunami (2005) demonstrated that these organisms possess highly developed olfactory learning capabilities, which are characterized by fast acquisition, long retention and easy memory recall. Additional studies showed that this form of learning was established early during the evolution of hemimetabolous insects, representing a valid tool for new insights into the evolution of neuronal systems subserving olfactory learning in both vertebrates and invertebrates (Taylor et al., 2020).

Not least, insects can be used as models for investigating the effects of herbicides on cognitive functions (Aloizou et al., 2020). One of the most widespread herbicides in the world is glyphosate N-(phosphonomethyl)-glycine (Ait-Bali et al., 2020). A decline in learning and navigation abilities has been observed in honeybees fed with concentrations of glyphosate similar to those found in the environment (Bara et al., 2014). Moreover, in the mosquitoes Aedes aegypti and Aedes albopictus, exposure to this herbicide altered larval development time and sex ratio, as well as the expression of genes conferring resistance to insecticides (Bara et al., 2014). In 2018, Balgan, Lazzari and Guerreri showed that the exposure of A. aegypti larvae to a field-realistic dose of glyphosate had deleterious effects on habituation learning (Baglan et al., 2018). This study opened the way for future ecotoxicological studies using mosquito larvae as a bio-indicator to evaluate the impact of herbicides, pollutants and chemical compounds on cognition.

Further examples of cognitive phenomena seen in insects include numerical cognition (Pahl et al., 2013), categorization of stimuli (Benard et al., 2006), cognitive/behavioral flexibility (Loukola et al., 2017; see Glossary), social learning and cultural transmission (Alem et al., 2016).

Planarians have recently become important model organisms in developmental and regenerative biology (Brown and Pearson, 2017). Their remarkable regenerative capacities – driven by an adult stem cell population – make them valid tools for investigating the molecular mechanisms behind neural repair and patterning. Furthermore, because of their rich behavioral repertoires and learning abilities, planarians are a potential tool for elucidating the dynamics of memory during brain regeneration. In 2013, Shomrat and Levin developed a system for investigating the dynamics of memory in a regenerating planarian's nervous system. For this purpose, they developed a computerized behavioral protocol to train flatworms in an environmental familiarization paradigm. As memory persisted for at least 14 days – a sufficiently long time for the nervous system to regenerate – they also demonstrated that trained, decapitated planarians exhibit memory retrieval after regenerating a new head (Shomrat and Levin, 2013). This study not only revealed LTM in planarians but also its persistence through head regeneration. The high tractability of this model system may shed light on the interface between body patterning and stored memories. Furthermore, future studies on these organisms may provide a great contribution to improving stem cell-derived treatments of degenerative brain disorders in human adults.

The studies on learning and memory in invertebrates find applications in our everyday life. We all live in a stressful world. Research over the last 40 years defined stress and the hormones and neurotransmitters released during and after a stressful event as major modulators of human learning and memory (Joëls et al., 2006; Vogel and Schwabe, 2016). Here, we define stress as a state that requires physiological and/or behavioral readjustment or modification to maintain the well-being of the organism (Lukowiak et al., 2014). The importance of invertebrate studies on stress relies on the fact that consistently with vertebrates, stress can alter adaptive behaviors, thereby either enhancing or diminishing learning and memory formation and/or recall.

As exemplified by the Yerkes–Dodson Law, too much or too little stress impedes LTM formation, while ‘just the right amount’ of stress enhances LTM (Teigen, 1994). Because LTM formation requires ‘neuronal cost’ (in terms of altered gene activity and new protein synthesis), organisms invest energy only for ‘relevant’ events. This relevancy is in part determined by the level of stress perceived. However, the same stimulus may be perceived as a stressor for one organism but not for another, or only at certain times and not at others in the same organism. Furthermore, populations or strains within a species may differ in their perception of, or response to, environmental stressors, showing – in turn – different adaptive behaviors and memory-forming potential (Lukowiak et al., 2014; Rivi et al., 2022f). Because stress has a broad definition, different stressors may have different biological consequences and, sometimes, opposite effects: some stressors block LTM formation, whereas others enhance it. This complex scenario is further complicated when multiple stressors are encountered. Although the exposure to different stressors may result in the same behavioral memory phenotype (i.e. memory enhancement or impairment), it is difficult to predict what the outcome will be regarding memory formation when a combination of stressors interact (Dalesman et al., 2013; Lukowiak et al., 2014). Thus, how combinations of stressors act is an emergent (i.e. basically unpredictable) property of how organisms perceive the stressors (Dalesman et al., 2013).

Given the complexity of vertebrate brains and behaviors, it is not too surprising that studies conducted in these models have led to sometimes contradictory results. Thus, studies using invertebrates may help researchers to decipher how stress affects memory at the behavioral, neuronal and molecular levels (Aonuma et al., 2018; Batabyal et al., 2022; Dalesman et al., 2013; Ito et al., 2015, 2017; Neckameyer and Nieto-Romero, 2015; Ottaviani and Franceschi, 1996; Rivi et al., 2022b; Soravia et al., 2021). For example, it has been demonstrated that different stress states resulting from different durations of food deprivation alter the ability of L. stagnalis to exhibit LTM (Ito et al., 2015). In particular, snails deprived of food for 1 day (a modest level of stress) before aversive classical conditioning (see Glossary) show optimal conditioned taste aversion (CTA; see Glossary) and LTM, whereas those starved for 5 days (high level of stress) before training do not show the memory phenotype. This is because severe food deprivation blocks the snails’ ability to express memory, which is formed but overpowered by severe hunger (Ito et al., 2015). This study demonstrated that CTA-LTM is both dependent on the level of stress and the context in which memory is formed. In fact, CTA-LTM memory expression occurs only if severely food-deprived snails are given ad libitum access to food for 7 days after training and are tested for memory recall after 1 day of starvation, which recreates the context in which they were trained (Ito et al., 2017).

Stress not only affects learning and memory but is also strongly associated with emotions. The link between cognitive function, stress and emotional states in humans has been largely demonstrated (Tyng et al., 2017). However, invertebrate studies in this field are only in their very early stages. Strange as it may seem, invertebrates exhibit various cognitive, behavioral and physiological traits that indicate internal states evocative of what we consider emotions. Emotions are defined as transient central states triggered by environmental stimuli resulting from an integration of subjective experiences, cognitive evaluation, behavior, neurophysiology and motivation (Anderson and Adolphs, 2014). For example, fear is a motivational state aroused by specific stimuli that give rise to defensive behavior or escape, whereas anxiety is a fear-related negative emotion, induced by a threat/aversive stressor to wellbeing or survival, either actual or potential (Steimer, 2002). In 2014, by using a sub-aquatic dark-light plus maze (a modified version of the famous plus-maze apparatus used for studying anxiety in rodents, consisting of light and dark arms), Fossat and coworkers studied stress-induced anxiety-related behaviors in crayfish (Procambarus clarkii). The exposure of animals to electric shocks before the maze test severely decreased exploratory behavior and increased light avoidance, a reaction strikingly similar to that observed in vertebrates (Fossat et al., 2014). These results not only emphasize the ability of an invertebrate model to exhibit a state that is like a mammalian emotion but may also have implications for learning and memory research. In fact, fear, anxiety and stress are inter-related in memory loss and cognitive decline in mammals (Sinoff and Werner, 2003).

Another important aspect that makes invertebrates valid models to study the intricate relationship between emotions, stress and cognition is that some physiological traits of stress and emotions described in vertebrates, such as increased heart and ventilation rate, also occur in invertebrates (Even et al., 2012; Ložek et al., 2019; Orr et al., 2007; Renwrantz and Spielvogel, 2011). For example, one of the CTA training procedures characterized in L. stagnalis used a conditional stimulus that elicits the feeding behavior (Benjamin and Kemenes, 2010) paired with an aversive unconditional stimulus that induces the whole-body withdrawal response and inhibits feeding (Ito et al., 1999; Sadamoto et al., 2010). Kita and colleagues (2011) demonstrated that after training and memory formation, the appetitive stimulus no longer elicited feeding, but increased the heart rate, inducing a response that was similar to fear in mammals (Kita et al., 2011; Steimer, 2002).

Furthermore, the high level of conservation of neurochemicals across taxa may be useful to the study of stress, emotions, learning and memory, and their interactions in invertebrates. The role of biogenic amines in the regulation of all these processes has been largely demonstrated in mammals (Purves et al., 2001). However, invertebrate nervous systems contain corresponding biogenic amines that are structurally and functionally similar to mammalian neurotransmitters, neuromodulators and hormones (D'Aniello et al., 2020). For example, training pond snails in the presence of their predator effluent enhances LTM formation and allows the memory to be recalled under a broader range of challenges (Dalesman et al., 2006; Forest et al., 2016). This stressor-mediated memory enhancement is prevented by exposure to the serotonin blockers mianserin and methysergide, suggesting a serotonergic modulation activated by risk perception, which results in enhanced memory formation (Forest et al., 2016). These data are consistent with previous studies in humans correlating increased serotonin levels with enhanced responses to anxiety-related stimuli and memory consolidation (Meneses, 2015; Wong et al., 2005).

These results may pave the way for future investigations aimed at unveiling the conserved core mechanisms of the intricate relationship between stress, emotion and cognitive function among a diversity of species and model animal systems.

As reported in this brief commentary, the use of invertebrates has significantly contributed to the characterization of the behavioral, neuronal and molecular mechanisms of memory formation, consolidation, reconsolidation and extinction. In addition, invertebrates can also be used to study the mechanisms through which some bioactive compounds (i.e. flavonoids), drugs (i.e. antidepressants, drugs of abuse, anti-inflammatory drugs, etc.), contexts and stressors affect (i.e. impair or enhance) memory.

We strongly believe that the use of invertebrates in applied neuroscience is only just beginning and that these organisms will provide a major contribution to answering the unsolved questions in neuroscience, revealing highly conserved characteristics between invertebrates and vertebrates, including humans.

The study of the connectome in insects and mollusks will provide important comparative information on how some neuronal connections have been maintained during the evolution despite the immense diversity in brain size and complexity across taxa. For the advancement of research in the field of neuroscience in the near future, more emphasis should be given to studies conducted using wild animals (Gibbs et al., 1997; Greif and Yovel, 2019; Kagan and Lukowiak, 2019; Masek et al., 2014; Swinton et al., 2021). In fact, as laboratory-inbred colonies tend to be subject to founder effects and genetic drift (see Glossary), they may not reflect the full range of behavioral responses and cognitive functions of natural populations (Brekke et al., 2018). We strongly believe – in accordance with numerous studies from rodents – that the human physiological and pathological processes related to memory and learning conditions are far closer to that of wild animals than to those of inbred ones (Festing, 1976; Rivi et al., 2022b; Tuttle et al., 2018).

Despite the extensive efforts in biomedical research, many aspects of neuro-aging are still not fully understood. Invertebrates such as mollusks (e.g. pond snails) and insects (e.g. flies) offer great advantages, including the accessibility and simplicity of their nervous system, which may contribute to the global understanding of neurophysiological processes underlying aging and memory decline at the genomic, neuronal and behavioral level (Fodor et al., 2021). For example, several relevant gero-protectors (i.e. molecules involved in protecting against aging) have been identified in Lymnaea, including gelsolin, presenilin, huntingtin, Parkinson's disease protein 7/protein deglycase DJ-1 and amyloid precursor protein (Fodor et al., 2021). These results strongly encourage future studies aimed at investigating the molecular, cellular and circuit mechanisms underlying the neurophysiological and neuropathological bases of aging and its effect on learning and memory abilities.

There has been a growing interest in determining the mechanisms underlying the individual cognitive abilities across organisms (Orr et al., 2009; Rivi et al., 2022b,c; Sunada et al., 2017). To date, only a few studies investigate the role (potentially) played by the gut microbiome as a driver of individual cognitive differences in natural populations of animals (Davidson et al., 2018). However, we believe that investigating the inter- and intra-specific variations in the gut microbiome will provide new insight into evolutionary and environmental mechanisms involved in cognitive functions across taxa. In fact, numerous studies demonstrated a major role of the microbiome composition in regulating neurotransmitter levels, the expression of neural receptors, synaptic plasticity and neurogenesis (Chen et al., 2021).

As previously reported, invertebrates such as Lymnaea can be extremely useful to decipher complex engram networks and provide a comprehensive map of engram circuits. The use of advanced tools for genome engineering is necessary for investigating the molecular mechanisms underpinnings of engram cells and their connections during memory formation and consolidation. Emerging in 2013, gene editing based on CRISPR-Cas9 technology represents a powerful strategy for efficiently manipulating key genes in multiple organisms, including invertebrates (Abe and Kuroda, 2019; Gratz et al., 2015; Jinek et al., 2012; Kohno et al., 2016; Martin et al., 2016; Sieber et al., 2021). Will the engram be found? Will it be discovered how to stop memory loss or keep certain memories unchanged over time? Will it be possible to erase negative experiences and their effects on memory? These questions will probably be found in the near future and invertebrates may be extremely useful to open new avenues of research and be part of the discoveries that will make us rethink many of our currently accepted beliefs.

We thank Stefan Galander and Charlotte Rutledge for the extraordinary editing work done on our manuscript, as well as the three anonymous reviewers for valuable comments.