Inverse Relationship between Basal Pacemaker Neuron Activity and Aversive Long-Term Memory Formation in Lymnaea stagnalis

Abstract

Learning and memory formation are essential physiological functions. While quiescent neurons have long been the focus of investigations into the mechanisms of memory formation, there is increasing evidence that spontaneously active neurons also play key roles in this process and possess distinct rules of activity-dependent plasticity. In this study, we used a well-defined aversive learning model of aerial respiration in the mollusk Lymnaea stagnalis (L. stagnalis) to study the role of basal firing activity of the respiratory pacemaker neuron Right Pedal Dorsal 1 (RPeD1) as a determinant of aversive long-term memory (LTM) formation. We investigated the relationship between basal aerial respiration behavior and RPeD1 firing activity, and examined aversive LTM formation and neuronal plasticity in animals exhibiting different basal aerial respiration behavior. We report that animals with higher basal aerial respiration behavior exhibited early responses to operant conditioning and better aversive LTM formation. Early behavioral response to the conditioning procedure was associated with biphasic enhancements in the membrane potential, spontaneous firing activity and gain of firing response, with an early phase spanning the first 2 h after conditioning and a late phase that is observed at 24 h. Taken together, we provide the first evidence suggesting that lower neuronal activity at the time of learning may be correlated with better memory formation in spontaneously active neurons. Our findings provide new insights into the diversity of cellular rules of plasticity underlying memory formation.

Keywords: Lymnaea stagnalis; aversive operant conditioning; basal neuronal activity; individual variations in memory formation; neuronal plasticity; spontaneously active neuron.

Figure 1

Aversive operant conditioning results in long-term memory (LTM) formation. (A) Time line of the conditioning paradigm. During the screening test, the frequency and total duration of pneumostome openings were observed over a 45 min period. Then, animals underwent two 45-min training sessions that were separated by an hour, during which trained animals received tactile stimuli contingent upon each attempted pneumostome opening and yoked controls received non-contingent stimuli. The frequency and total duration of pneumostome openings were observed again 24 h after conditioning during the memory test. Animals that underwent operant conditioning show significant reductions in both the total duration (B) and frequency (C) of pneumostome openings 24 h after conditioning as compared with the screening session, indicating LTM formation. No significant difference in either parameter is observed in yoked controls. In order to separate conditioned animals that do and do not form LTM, each animal was scored using the ratio of total duration of pneumostome opening at 24 h to those at screening (“D24/D0” ratio) to assess its behavioral response to the conditioning procedure. Animals that exhibit D24/D0 ratio <0.6 were classified as “LTM” and those with D24/D0 >0.8 were classified as “no-LTM”. Summary of total duration (D) and frequency (E) of pneumostome openings in yoked control, no-LTM and LTM animals before and after 24 h conditioning. Whereas LTM animals exhibit significant reductions in both the total duration and frequency of pneumostome openings at 24 h after conditioning, no-LTM animals did not. **p < 0.01, ***p < 0.001, ****p < 0.0001 as compared to screening test in LTM animals. ^###p < 0.001, ^####p < 0.0001 as compared to screening test in no-LTM animals.

Figure 2

LTM is associated with enhanced RPeD1 firing activity. (A) Intracellular sharp electrode recordings of the RPeD1 were performed in isolated preparations of the central ganglia ring to measure its membrane potential, spontaneous firing frequency, gain of firing and input resistance as indices of neuronal firing properties. White arrow indicates the location of RPeD1. (B) The evoked activity step protocol. (C) Representative traces of RPeD1 firing upon current injection in no-LTM, LTM and yoke animals. Animals with LTM exhibit significantly depolarized membrane potential (D), increased spontaneous firing frequency (E), enhanced gain of firing (F) and reduced input resistance (G) as compared to no-LTM and yoked control animals, whereas the latter two groups do not differ from each other. *p < 0.05, **p < 0.01, as compared to yoked control. ^#p < 0.05, ^##p < 0.01 as compared to no-LTM.

Figure 3

Relationship between aerial respiration behavior and RPeD1 firing activity in naïve animals. (A) Intracellular sharp electrode recordings of the RPeD1 were performed in isolated preparations of the central ring ganglia. Correlations between total duration of pneumostome openings and membrane potential (A), spontaneous firing frequency (B), gain of firing (C) and input resistance (D).

Figure 4

Endogenous differences in basal aerial respiration behavior is correlated with different responses to aversive operant conditioning. (A) Quartile analysis of behavioral data from trained animals demonstrated that animals whose basal total duration of aerial respiration behavior was in the upper two quartile exhibited significant reductions in the number of conditioning stimuli received during the second training session (T2) as compared to the first (T1), whereas animals in the lower two quartile did not. Retrospective analysis using the average ratio of number of conditioning stimuli received during T2 to those during T1 (“T2/T1 ratio”) in animals in the upper two quartiles, ~0.6, and lower two quartiles, ~0.8, as the criteria showed that animals that exhibited T2/T1 ratio <0.6 showed reduced total duration (B) and frequency (C) of pneumostome opening at 24 h after conditioning, whereas those that exhibited T2/T1 ratio >0.8 did not. ^##p < 0.01, ^###p < 0.001 as compared to T1. ***p < 0.001, ****p < 0.0001 as compared to screening test. ^#p < 0.05 as compared to screening test level in T2/T1 ratio >0.8 animals.

Figure 5

RPeD1 firing activity is enhanced in a time-dependent manner during the first 24 h after conditioning in animals that form LTM. Central ring ganglia of no-LTM and LTM animals were extracted at various time points during the first 24 h after conditioning. Intracellular sharp electrode recording of the RPeD1 was used to measure its membrane potential (A), spontaneous firing frequency (B), gain of firing (C) and input resistance (D). As compared to no-LTM animals (1 h: n = 5; 2 h: n = 6; 3 h: n = 8; 4–6 h: n = 8; 6–8 h: n = 6; 8–10 h: n = 6; 24 h: n = 8), RPeD1 of animals with LTM (1 h: n = 6; 2 h: n = 6; 3 h: n = 6; 4–6 h: n = 6; 6–8 h: n = 6; 8–10 h: n = 6; 24 h: n = 8) exhibit two temporally discontinuous phases of enhancement in membrane potential, spontaneous firing frequency and gain of firing: an early phase that decays within 3 h and a late phase that persists by 24 h. Input resistance in LTM animals is lower than that of no-LTM animals at 8–10 and 24 h. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 6

Proposed model describing the role of basal RPeD1 firing activity in aversive LTM formation. (A) In naïve animals, differences in endogenous expression and/or activity level of several ion channels may underlie individual variations in basal RPeD1 rhythmic firing activity, which is negatively correlated with basal aerial respiration behavior. Animals that form aversive LTM exhibit lower basal RPeD1 firing activity and higher basal aerial respiration behavior than animals that do not form aversive LTM. (B) During training, animals with high- and low-firing RPeD1s exhibit different responses to the conditioning stimuli, as reflected in the different number of conditioning stimuli delivered in T2. Application of a conditioning stimulus (blue inverted triangle) to the open pneumostome results in a transient inhibition of RPeD1 firing activity. The inhibitory input strongly suppresses spike firing in a RPeD1 with lower basal firing activity, resulting in a reduction in intracellular Ca²⁺ concentration that has been shown to trigger inhibition of CaMKII () and K_Ca channel activity () (potential modifications signified by yellow circle), and potentiation of firing activity (Nelson et al., 2003, 2005) (C). In contrast, the conditioning stimulus does not significantly induce changes in the spike firing and intracellular Ca²⁺ concentration in a RPeD1 with higher basal firing activity, such that cellular changes underlying aversive LTM formation do not occur (C). (C) In animals exhibiting low basal RPeD1 firing, the training-induced enhancement in RPeD1 rhythmic firing activity observed at 1 h after training could induce increases in intracellular Ca²⁺ concentration and consequently activity-dependent signaling pathways (), such as CREB (Guo et al., 2010), that trigger aversive LTM formation. At 24 h after training, de novo synthesis of plasticity products (), such as voltage-gated Ca²⁺ and non-selective Na⁺ leak channel (NALCN) leak channels, may serve to maintain enhanced RPeD1 firing activity and aversive LTM.