Insulin and insulin-like peptides (ILP) help to maintain glucose homeostasis, whereas insulin-like growth factor (IGF) promotes the growth and differentiation of cells in both vertebrates and invertebrates. It is sometimes difficult to distinguish between ILP and IGF in invertebrates, however, because in some cases ILP has the same function as IGF. In the present review, therefore, we refer to these peptides as ILP/IGF signaling (IIS) in invertebrates, and discuss the role of IIS in memory formation after classical conditioning in invertebrates. In the arthropod Drosophila melanogaster, IIS is involved in aversive olfactory memory, and in the nematode Caenorhabditis elegans, IIS controls appetitive/aversive response to NaCl depending on the duration of starvation. In the mollusk Lymnaea stagnalis, IIS has a critical role in conditioned taste aversion. Insulin in mammals is also known to play an important role in cognitive function, and many studies in humans have focused on insulin as a potential treatment for Alzheimer’s disease. Although analyses of tissue and cellular levels have progressed in mammals, the molecular mechanisms, such as transcriptional and translational levels, of IIS function in cognition have been far advanced in studies using invertebrates. We anticipate that the present review will help to pave the way for studying the effects of insulin, ILPs, and IGFs in cognitive function across phyla.

Adiponectin enhances insulin sensitivity, which improves cognition in mammals. How adiponectin affects the mechanism’s underlying cognition, however, remains unknown. We hypothesized that experiments using the pond snail Lymnaea stagnalis, which has long been used in learning and memory studies and in which the function of insulin-like peptides affect learning and memory, could clarify the basic mechanisms by which adiponectin affects cognition. We first identified putative molecules of adiponectin and its receptor in Lymnaea. We then examined their distribution in the central nervous system and changes in their expression levels when hemolymph glucose concentrations were intentionally decreased by food deprivation. We also applied an operant conditioning protocol of escape behavior to Lymnaea and examined how the expression levels of adiponectin and its receptor changed after the conditioned behavior was established. The results demonstrate that adiponectin and adiponectin’s receptor expression levels were increased in association with a reduced concentration of hemolymph glucose and that expression levels of both adiponectin and insulin-like peptide receptors were increased after the conditioning behavior was established. Thus, the involvement of the adiponectin-signaling cascade in learning and memory in Lymnaea was suggested to occur via changes in the glucose concentrations and the activation of insulin.

Quantitative real-time PCR (qPCR) is a powerful method for measuring nucleic acid levels and quantifying mRNA levels, even in single cells. In the present study, we compared the results of single-cell qPCR obtained by different quantification methods (relative and absolute) and different reverse transcription methods. In the experiments, we focused on the cerebral giant cell (CGC), a key neuron required for the acquisition of conditioned taste aversion in the pond snail Lymnaea stagnalis, and examined changes in the mRNA levels of 3 memory-related genes, cAMP-response element binding proteins (LymCREB1 and LymCREB2) and CREB-binding protein (LymCBP), during memory formation. The results obtained by relative quantification showed similar patterns for the 3 genes. For absolute quantification, reverse transcription was performed using 2 different methods: a mixture of oligo d(T) primers and random primers (RT method 1); and gene-specific primers (RT method 2). These methods yielded different results and did not show consistent changes related to conditioning. The mRNA levels in the samples prepared by RT method 2 were up to 3.3 times higher than those in samples prepared by RT method 1. These results suggest that for qPCR of single neurons, the efficacy and validity do not differ between relative and absolute quantification methods, but the reverse transcription step critically influences the results of mRNA quantification.

Some evidence suggests that oxytocin, which is a neuropeptide conventionally thought to be synthesized in the hypothalamus and released by the posterior pituitary, is generated in peripheral keratinocytes, but the details are lacking and the mRNA analysis is further required. Oxytocin and neurophysin I are generated together as cleavage products after splitting the precursor molecule, preprooxyphysin. To confirm that oxytocin and neurophysin I are also generated in the peripheral keratinocytes, it must first be clarified that these molecules contained in peripheral keratinocytes did not originate in the posterior pituitary gland and then the expression of oxytocin and neurophysin I mRNAs must be established in keratinocytes. Therefore, we attempted to quantify preprooxyphysin mRNA in keratinocytes using various primers. Using real-time PCR, we observed that the mRNAs of both oxytocin and neurophysin I were located in keratinocytes. However, the mRNA amounts of oxytocin, neurophysin I, and preprooxyphysin were too small to confirm their co-existence in keratinocytes. Thus, we had to further determine whether the PCR-amplified sequence was identical to preprooxyphysin. The PCR products analyzed by DNA sequencing were identical to preprooxyphysin, finally determining the co-existence of both oxytocin and neurophysin I mRNAs in keratinocytes. In addition, the immunocytochemical experiments showed that oxytocin and neurophysin I proteins were located in keratinocytes. These results of the present study provided further support indicating that oxytocin and neurophysin I are generated in peripheral keratinocytes.

The pond snail Lymnaea stagnalis is capable of learning by both classical conditioning and operant conditioning. Although operant conditioning related to escape behavior with punishment has been examined by some research groups, the molecular mechanisms are not known. In the present study, we examined changes in the expression levels of cAMP-response element binding protein 1 (CREB1), CREB2, CREB-binding protein (CBP), and monoamine oxidase (MAO) in the Lymnaea central nervous system (CNS) using real-time PCR following operant conditioning of escape behavior. CREB1 and CREB2 are transcription factors involved in long-term memory in Lymnaea; CBP is a coactivator with CREB1; and MAO is a degrading enzyme for monoamines (e.g., serotonin) with important roles in learning and memory in Lymnaea. In operant conditioning, the punishment cohort, in which snails escaping from the container encountered aversive KCl, exhibited significantly fewer escape attempts than the control cohort, in which snails escaping from the container encountered distilled water, during both the training and memory test periods. After the operant conditioning, CREB1 and CREB2 were upregulated, and the ratio of CREB1/CREB2 was also increased, suggesting that the operant conditioning of escape behavior involves these factors. MAO was also upregulated, suggesting that the content of monoamines such as serotonin in the CNS decreased. The upregulated genes identified in the present study will help to further elucidate learning and memory mechanisms in Lymnaea.