ABSTRACT
Social interactions play an important role in learning and memory. There is great variability in the literature regarding the effects of social isolation on cognition. Here, we investigated how memory formation was affected when Lymnaea stagnalis, our model system, were socially isolated at three different time periods: before, during or after the configural learning training procedure. Each group of snails underwent configural learning where we recorded and compared their feeding behaviour before and after the pairing of an appetitive food stimulus with predator kairomones (i.e. the training procedure). We found that isolating snails before the training procedure had no effect on their learning and memory. However, when snails were isolated either during the training procedure or immediately after the training procedure, they no longer formed memory. These data provide further insight into how isolation impacts cognitive functioning in the context of higher-order learning.
INTRODUCTION
Social interactions are important for facilitating new learning and modifying pre-existing memories, as well as helping buffer stress or the effects of negative memories. In both mammals and invertebrates, social isolation significantly alters behaviour and physiology. When rodents are socially isolated, they exhibit cognitive and emotional abnormalities (Zhao et al., 2009) and changes in open field behaviour (Krohn et al., 2006). In the same vein, invertebrates are also affected by social isolation (Sokolowski, 2010). Socially isolated Caenorhabditis elegans experience impaired development, reduced formation of neural connections and a decline in behavioural responsiveness (Rose et al., 2005).
Social isolation has also been associated with cognitive decline via stress-associated changes in the nervous system of invertebrates. For example, socially isolating Drosophila melanogaster results in reduced fibre density in mushroom bodies (Technau, 2007), which play an important role in learning and memory in insects (Fahrbach, 2006). Similarly, socially isolating the bee Apis mellifera ligustica results in a reduction in mushroom body volume (Maleszka et al., 2009). Additionally, marine molluscs Aplysia fasciata that are socially isolated during or shortly after behavioural training experience a block in long-term memory (LTM) formation (Schwarz et al., 1998).
The great pond snail, Lymnaea stagnalis, is an established model system for studying learning and memory because of its relatively simple nervous system and readily measurable tractable behaviours (Benjamin et al., 2000). Moreover, in L. stagnalis, memory formation is malleable and can be either enhanced or blocked by a variety of ecologically relevant stressors (Lukowiak et al., 2010). Previous studies from our lab demonstrate that L. stagnalis are affected by changes in their social environment. Crowding of same-strain snails was found to be a stressor causing an obstruction of LTM formation (De Caigny and Lukowiak, 2008; Dodd et al., 2018). Interestingly, when snails were crowded with a different species of snail or different strains/populations of L. stagnalis, crowding no longer obstructed LTM formation (Dodd et al., 2018). This finding is consistent with the notion that L. stagnalis can differentiate between different strains/populations and are impacted by certain social situations. Additionally, there are studies that show that social isolation impacts snail behaviours. The first of these showed that social isolation results in significant changes in reproductive behaviour (Koene et al., 2009). Thus, prolonged isolation of L. stagnalis during development leads to delayed reproduction and increased size at the onset of reproduction (Koene et al., 2009). Additionally, maintaining snails in social isolation before operant conditioning of aerial respiration both enhanced LTM formation under certain conditions and blocked LTM formation under others (Dalesman and Lukowiak, 2011). This suggests that the effect of social isolation on cognitive function in L. stagnalis depends on the unique environmental context in which it is experienced. Here, we wanted to further explore the importance of a social setting for learning and memory in L. stagnalis. Specifically, we wanted to study whether the presence of other snails was necessary for snails to successfully undergo configural learning, a more complicated form of associative learning that occurs in the snails' natural environment and is important for their survival (Swinton et al., 2019; Kagan and Lukowiak, 2019). In brief, configural learning is a form of higher-order learning where the organism forms an association between two stimuli experienced together that is different from the simple sum of their components (Pearce, 2002; Giurfa, 2003).
Lymnaea stagnalis form configural learning when they experience food (e.g. carrot odour, CO) together with a predator scent (crayfish effluent, CE) (Swinton et al., 2019). Following configural learning (i.e. CO+CE), the carrot odour no longer elicits a feeding response, and instead elicits a fear state resulting in a suppressed feeding response. The CO+CE training procedure that was used in the previous configural learning experiments for L. stagnalis involves a cohort of snails being housed together before and after the CO+CE pairing and snails all being in the same beaker during the CO+CE pairing. Thus, throughout the entire configural learning procedure, snails are always together. To test whether social isolation would impact snails' ability to successfully undergo configural learning and form memory, we isolated snails before, during or following (i.e. during the memory consolidation phase) the configural learning procedure. We hypothesized that if snails were socially isolated prior to the configural learning procedure, memory formation would not be affected. However, if snails were socially isolated during or immediately after the configural learning procedure, they would not form memory. We know that stressors alter learning and memory in the pond snail, with too much or too little stress impeding memory formation (Lukowiak et al., 2014). For example, when experienced as a single stressor, CE enhances memory formation for operant conditioning in L. stagnalis. However, when the CE stressor is combined with the stress from crowding, snails are no longer able to form memory (Lukowiak et al., 2014). As such, it could be the case that combining the stress of being isolated with the stress of a predator scent during the critical periods of memory formation could result in too much stress, impeding snails' ability to learn and form memory.
MATERIALS AND METHODS
Snails and animal maintenance
We used an inbred laboratory strain (W-strain) of Lymnaea stagnalis (Linnaeus 1758) maintained at the University of Calgary Biology department. W-strain snails are considered ‘predator experienced’. That is, even though they have been ‘born and bred’ in Calgary for many years in an environment in which they never encounter crayfish as a predator, they do respond to the scent of a crayfish (CE) with anti-predatory behaviours, including enhanced memory formation (Orr et al., 2007; Orr and Lukowiak, 2008). Snails were housed in artificial pond water (0.25 g l−1 of Instant Ocean in deionized water, Spectrum Brands, Madison, WI, USA) supplemented with CaCO3 to ensure calcium concentrations remain above 50 mg (Dalesman and Lukowiak, 2011). We maintained snails at 20±1°C on a 16 h:8 h light:dark cycle. Snails fed on romaine lettuce ad libitum.
Rasping behaviour
A rasp in L. stagnalis is a rhythmic motor behaviour in which repeated movements of the radulae scrape the surface of a substrate, leading to the ingestion of food (Ito et al., 2013). To observe rasping behaviour, snails were placed in a 14 cm Petri dish positioned above a mirror with enough carrot slurry for them to be partially submerged. Snails were given approximately 2 min to acclimate before we began counting rasps. After acclimation, each snail was observed for 2 min, and the number of rasps were counted. We then calculated the average number of rasps per minute and graphed these values.
Configural learning
In this associative learning paradigm, snails were exposed simultaneously to the carrot slurry (i.e. CO) and crayfish effluent (CE) for 45 min. This is referred to as the configural learning procedure. Carrot slurry is an appetitive food stimulus that elicits robust feeding behaviour. The carrot slurry was made by blending two peeled medium-sized (60–70 mg) commercially obtained organic carrots along with approximately 500 ml of pond water. Following blending and repeated straining (at least 3 times) of the mixture, a liquid carrot–pond water slurry was obtained which contained no visible pieces of carrot.
Crayfish are natural predators of Lymnaea (Orr et al., 2007). In this study we used an Orconectes virilis crayfish that was commercially obtained. The crayfish was housed in a 70 l aquarium of artificial pond water and was fed fish pellets, lettuce and snails. We utilized the water from the crayfish tank as the CE (Orr and Lukowiak, 2008).
For the experiments described below, we first recorded each snail's rasping behaviour in the carrot slurry. The snails were then returned to their home aquaria. The following day (i.e. 18 h later), snails were trained using the configural learning training procedure (i.e. CO+CE) for 45 min and then were returned to their home aquaria. Two hours after the configural learning training procedure, we again recorded snails' rasping behaviour in the carrot slurry only.
Social isolation
In this study, we explored the effect of socially isolating snails at three different time points in the configural learning procedure: (1) for 18 h before the configural learning training procedure; (2) during the configural learning training procedure; or (3) in the 3 h immediately after the configural learning training procedure. When snails were socially isolated, they were housed in their own small aquaria with approximately 300 ml of pond water and were fed lettuce ad libitum. It is important to note that snails were not socially isolated in any of the experiments shown here before their first exposure to carrot.
Statistical analyses
All data were analysed and graphed using GraphPad Prism v.9.00e for MAC® (GraphPad Software, Inc., La Jolla, CA, USA). All datasets were determined to be normally distributed. Thus, a paired t-test was performed to examine the difference in the number of rasps in CO before and after CO+CE pairing. Differences were considered significant if P<0.05. Data are presented as means±s.e.m.
RESULTS
Effects on LTM formation following social isolation at different time points
We first performed a control experiment (Fig. 1) with no social isolation at any time point to demonstrate successful configural learning. A paired t-test indicated that snails rasped significantly fewer times in carrot slurry following (CO post) the configural learning training procedure (CO+CE) compared with before pairing (CO pre; paired t-test, t=3.384, d.f.=10, P=0.0070, n=11). Next, we tested whether social isolation for 18 h before the configural learning training procedure would have an impact on memory formation (Fig. 2). As can be seen, we obtained similar data to those in the control experiment shown in Fig. 1. That is, snails rasped significantly less after (CO post) versus before (CO pre) the configural learning training procedure (CO+CE) (paired t-test, t=6.175, d.f.=9, P=0.0002, n=10). Thus, social isolation before the configural learning procedure did not result in a blockage of memory formation.
We then isolated snails during the 45 min configural learning training procedure and studied the effect of that isolation on memory formation (Fig. 3). When snails were isolated during the configural learning training procedure (CO+CE), training no longer caused a significant difference in rasps between CO pre and CO post (paired t-test, t=1.096, d.f.=9, P=0.3016, n=10). Thus, when snails are socially isolated during the configural learning training procedure, they do not form an associative memory.
Finally, we examined the impact that social isolation for 3 h immediately following the configural learning training procedure has on memory formation (Fig. 4). Here, as in Fig. 3, snails again showed no significant difference in rasping behaviour in CO pre versus CO post (paired t-test, t=1.164, d.f.=9, P=0.2743, n=10). Thus, social isolation immediately following the configural learning training procedure also inhibited memory formation.
DISCUSSION
Here, in our Lymnaea model system, we investigated the effect of social isolation on the ability to form a configural learning memory that has been previously demonstrated in these snails (Swinton et al., 2019). We hypothesized that isolating snails for 18 h prior to the configural learning procedure (i.e. CO+CE) would have no effect on their ability to form memory. The data we obtained here are consistent with that hypothesis and demonstrate that the memory formed in socially isolated snails before training was not different from that in control experiments here (i.e. Figs 1 and 2) and previously published results (Swinton et al., 2019). These findings are consistent with the data obtained with operant conditioning of aerial respiratory behaviour (Dalesman and Lukowiak, 2011). In that study, researchers found that in the presence of predator kairomones, maintaining snails in isolation prior to training had no effect on their ability to learn and form LTM. As shown here, while the combination of predator kairomones (i.e. CE) and social isolation may cause stress, the period in which the stressors are perceived – prior to the configural learning training procedure – is not a critical time point for memory formation. Rather, as the subsequent experiments showed, the perception of the social isolation stress during the configural learning training procedure or immediately after that training altered the ability of the snails to form a configural learning memory.
Isolating snails either during or after the configural learning training procedure blocked their ability to form memory. These data are consistent with the data obtained in a related gastropod mollusc, Aplysia fasciata (Schwarz and Susswein, 1992). There, it was shown that in the absence of a conspecific experience, learning of a feeding task was inhibited. However, the absence of conspecifics in those experiments also altered other aspects of homeostatic feeding behaviours. Thus, those authors, concluded that interference with learning caused by social isolation was due to an inhibition of feeding and not because of a direct effect on learning. Here, we know that this is not the case as snails that were isolated for 18 h before the configural learning procedure performed normally in regard to both feeding and configural learning.
Interestingly, our findings at one level contrast with those from our previous study (Dalesman and Lukowiak, 2011) which demonstrated that when isolated snails were operantly conditioned (aerial respiration) in CE, they still exhibited enhanced LTM formation. That is, social isolation did not alter the enhancing ability that CE had on LTM formation. However, this was suggested to be because the effects of social isolation on cognitive function vary and are highly dependent on the context in which social isolation is experienced (Dalesman and Lukowiak, 2011). That is, social isolation had: (1) no discernible effect, (2) an enhancing effect, or (3) a blockade effect on memory formation depending on the presence or absence of other memory-altering stressors. Thus, social isolation could block LTM formation when snails were both exposed to CE and trained in low calcium conditions. This result was explained by hypothesizing that under those conditions (i.e. CE, low external calcium and social isolation), snails were experiencing too much stress. According to the Yerkes–Dodson–Hebb ‘law’, there is an optimum level of stress, below or above which attention is not paid to training, impairing learning (Yerkes and Dodson, 1908; Ito et al., 2015). This ‘law’ has been previously demonstrated in Lymnaea, where too little or too much of a KCl bath stressor experienced before training resulted in reduced memory formation (Martens et al., 2007; Ito et al., 2015).
In our present study, social isolation affected memory formation in the presence of only one environmental stressor, predator kairomones (i.e. CE). Thus, our results differ from those previously obtained (Dalesman and Lukowiak, 2011) and this may be for several reasons. Firstly, our methods of isolation differed. In our previous study (Dalesman and Lukowiak, 2011), snails were isolated by placing them in individual perforated containers within a single, larger aquarium; thus, waterborne chemical signals could travel between animals. This creates the possibility that snails were able to use those waterborne chemicals to detect the presence of nearby conspecifics in the immediate vicinity (thereby making them ‘not isolated’), hence isolation not altering the ability of snails to form LTM in the presence of predator kairomones alone (Dalesman and Lukowiak, 2011). Secondly, the difference in learning procedures may account for the different results. Our previous study (Dalesman and Lukowiak, 2011) utilized operant conditioning of aerial respiration, which is considered an elemental or ‘lower-order’ form of associative learning (Swinton et al., 2019). In the current study, we utilized configural learning, a higher-order form of associative learning (Giurfa, 2003). Interested readers who wish to learn more about configural learning and how configural learning as shown in Lymnaea compares to that in mammalian organisms are directed to papers by Pearce (2002) and Sutherland and Rudy (1989). Higher-order conditioning requires that the animal make meaningful comparisons between current sensory stimuli and its representation of previous sensory experiences, making it more ‘cognitive’ compared with lower-order learning (Sahley et al., 1990; Hawkins et al., 1998; Devaud et al., 2015). Therefore, it may be the case that higher-order learning is more sensitive to the effects of stress and the optimum level of stress required for successful learning may be less than that required for lower-order conditioning. Also, the creation of a landscape of fear (Laundre et al., 2010) may be similar in the two cases but the consequences of that would differ because of the higher level of cognition required for configural learning compared with operant conditioning. These ideas are supported by previous studies which show differing effects of the same stressor on different types of learning. For example, in food deprivation studies, one day of food deprivation resulted in enhanced memory for snails trained with conditioned taste aversion (Ito et al., 2017) but not for those trained with operant conditioning (Kagan et al., 2023). In contrast, longer lengths of food deprivation inhibited memory for snails trained with conditioned taste aversion (Ito et al., 2017) but enhanced memory for those trained with operant conditioning (Kagan et al., 2023). Thus, the optimum level of stress required for successful learning and memory formation seems to vary based on the type of learning procedure used.
Previous literature has shown varying effects of isolation on cognitive functioning in a wide range of species (Cacioppo and Hawkley, 2009). In L. stagnalis, there has been evidence to support neutral, negative and enhanced effects of isolation on memory formation following operant conditioning depending on the context of the environmental stress experienced (Dalesman and Lukowiak, 2011). In the current study, we demonstrated that isolation before the configural learning training procedure has neutral effects on memory formation, while isolation during or after that training has negative effects on memory formation. Thus, the results of this study further support the conclusion that social isolation has varying effects on cognition. How stressors experienced together interact to alter cognitive ability is thus complicated and not easily predictable.
Our present study is limited in that we only explored the effects of isolation on one form of higher-order cognition. In future studies, we plan to investigate the effects of isolation on other more complicated forms of learning such as the Garcia effect to better understand when social isolation impacts learning and memory. The Garcia effect is another form of higher-order learning in which animals acquire a taste aversion after experiencing a novel food substance and a visceral sickness hours following the taste of the novel food (Garcia et al., 1995; Rivi et al., 2021). Additionally, in future research we will consider investigating the transcriptional effects induced by isolation so that we can make more direct connections regarding the effect that isolation has on the mechanisms underlying learning and memory. Overall, the results of this study and the future research questions it has inspired will provide further insight into the varied effects that isolation has on cognition.
Acknowledgements
We would like to thank Petronella Hermann from the University of Calgary for providing the experimental animals used in our experiments.
Footnotes
Author contributions
Conceptualization: D.K., A.B., K.L.; Methodology: K.L.; Validation: K.L.; Formal analysis: J.H., A.B.; Investigation: J.H., A.B.; Resources: K.L.; Writing - original draft: D.K., K.L.; Writing - review & editing: D.K., K.L.; Supervision: K.L.; Project administration: K.L.; Funding acquisition: K.L.
Funding
This research was funded by the Natural Sciences and Engineering Research Council of Canada (NSERC-227993-2019) to K.L. This source of funding was not involved in the study design, data collection, analysis, interpretation, writing of the report, or in the decision to submit the paper for publication.
Data availability
All relevant data can be found within the article.
References
Competing interests
The authors declare no competing or financial interests.