Sensory Plasticity Caused by Up-down Regulation Encodes the Information of Short-term Learning and Memory

doi:10.21203/rs.3.rs-4917555/v1

Download PDF

Article

Sensory Plasticity Caused by Up-down Regulation Encodes the Information of Short-term Learning and Memory

https://doi.org/10.21203/rs.3.rs-4917555/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Learning and memory are essential for animals’ well-being and surviving. The underlying mechanisms are a major task of neuroscience studies. In this study, we identified a circuit consisting ASER, RIC, RIS, and AIY, required for short-term salt chemotaxis learning (SCL) in C. elegans. ASER NaCl sensation possesses ON- and OFF-responses and is remodeled by salt conditioning. RIC integrates sensory information of NaCl stimulation and food deprivation and generates a suppression of its NaCl calcium response by salt conditioning. RIC plasticity combining with interaction between ASER and RIS generates ASER sensory plasticity that is required for learning and memory recalling. We further identify the signaling pathways between neurons in the circuit: tyramine/TYRA-2 and octopamine/OCTR-1 signaling mediate neurohumoral regulation of AIY and ASER by RIC; FLP-14/FRPR-10 and PDF-2/PDFR-1 signalings mediate the transmission of humoral regulation from ASER to RIS and the feedback from RIS to ASER, respectively. Thus, ASER sensory plasticity encodes the information of the short-term SCL, which can facilitate animal adaptation to dynamic environments.

Biological sciences/Neuroscience/Neural circuits

Biological sciences/Neuroscience/Learning and memory/Short-term memory

learning and memory

sensory plasticity

neuropeptide receptors

C. elegans.

Learning and memory (L&M) are essential for animals’ well-being and surviving in complex and dynamic environments. Learning refers to the acquisition of new information or knowledge or a relatively permanent change in behavior as the result of experience. Memory refers to the process by which that knowledge is encoded, stored, and later retrieved. There are multiple forms of learning and memory. Each form of L&M has distinctive cognitive and computational properties and is supported by different brain systems. Memory can be classified into different modules along two dimensions: short-term, intermediate, and long-term memory by the time course of storage, and implicit and explicit memory by the nature of the information stored. Various modules of L&M involve different neural structures and different mechanisms. Various structures of the nervous system in different animals, particularly in the variety of forms or patterns of learning memory. The neural structures or brain regions and circuits involved in learning memory activities may differ in different species of animals. However, the basic molecular, cellular, and circuital mechanisms of L&M are highly conserved in the animal kingdom^1–3.

The study of learning and memory was a branch of philosophy for most of human history. Today, it is an important topic of neuroscience⁴. L&M and underlying mechanisms have been systematically and extensively studied. An enormous amount of experimental research is directed toward elucidating the neurobiological substrates that support many forms of learning and memory as well as those cognitive processes that utilize memory⁵. It is well accepted in neuroscience that all forms of L&M depend on structural and functional plasticity of the nervous system. L&M mechanisms are complex, involving sensation, neural signaling, and its plasticity^{2, 6, 7}, analytic integration of neural signals in neural circuits^{8, 9}, formation of new synapses or neural circuits, and changes in system-level activity^{10, 11}.

The mechanisms of nervous plasticity on levels of molecular reaction cascades, cells, neuronal circuits, and nervous system are one of major tasks of neuroscience studies¹². Neural circuits acquire new information by changing network properties on the level of specified neurons and their synaptic connections. It is the task of today’s neuroscience to unravel the complex hierarchies of interactions from molecular to systematic levels in solving the problem of predicting future behavior from experience in the past.

Both genes and the modalities of neural information integration are conserved among animal genera. Experimentally tractable animal models are generally used to study the relationships among the genes, proteins, and neural circuits underlying behaviors. Invertebrate nervous systems, because of their relative simplicity, offer significant advantages for multidisciplinary molecular, cellular, genetic, and behavioral investigations of the mechanisms underlying learning, memory formation, and memory retrieval¹². The nematode Caenorhabditis elegans (C. elegans) is a favored model for behavioral studies because of rapid life cycle, large number of progenies, ease of cultivation in a laboratory setting, a compact nervous system, a better-worked-out toolbox of molecular genetics and neuronal manipulations, and a wide range of behavioral tests has already been analyzed. C. elegans exhibits a great deal of plasticity and a broad range of learning and memory abilities, including that of mechanosensory, thermosensory, olfactory, gustatory, etc^12–15.

Short-term salt chemotaxis learning is essential for optimizing the chance of survival. Food-finding and nutrients uptake are essential for all animals’ survival. Therefore, animals are endowed with learning and memory for conditions of food availability. They learn to associate ambient conditions with availability of food, and they later pursue the conditions with abundant food or avoid the conditions with scarce or no food. C. elegans shows chemotaxis to various volatile and water-soluble chemicals^{16, 17}, which are often categorized as odor and taste substances by analogy with vertebrates. C. elegans shows chemotaxis to sodium chloride (NaCl), a chemical as an indicant for the availability of bacteria food. The direction and extent of the salt chemotaxis change depending on experience with salt and food. After association of NaCl with lack of food for a short period (10 min), C. elegans showed a dramatic reduction of chemotaxis to NaCl and eventually an avoidance of the salt (a negative chemotaxis against NaCl). This model of behavioral plasticity is termed as short-term salt chemotaxis learning (SCL) or gustatory aversive learning^{15, 18–20}.

Multiple classical neurotransmitters, neuromodulators, and neuropeptides are identified to engage in C. elegans SCL. Moreover, multiple neurons, including ASER, AIB alone and the combination of AIY and AIZ are needed for short-term SCL. ASER senses NaCl stimulation and may function as site of the memory of the sensory experience²¹.

Small molecule neurotransmitters or modulators are the main signaling substances that mediate signaling in the nervous system, and transmitter release and signaling system plasticity are the basis for learning and memory^{15, 22}. Among small molecule neurotransmitters and neuromodulators, and tyramine (TA) and octopamine (OA), mammalian homologs of epinephrine and norepinephrine, function as important modulators for short-term SCL^{23, 24}.

Neuropeptides are short peptides (3 ~ 100 amino acids) secreted by neurons through the regulatory secretion pathway that act on neurons and other tissue cells. They are important modulators and even neurotransmitters in the nervous system. Neuropeptides play very important roles in regulating nervous system activities, which modulate, among other things, emotions (fear and anxiety), appetite and obesity, sleep, alcohol dependence, learning and memory, and multiple modes of sensation and behavior in both humans and animals^25–31.

The C. elegans genome encodes at least 153 neuropeptide-encoding genes that give rise to over 300 predicted bioactive peptides, many of which have been biochemically confirmed^32–34. The neuropeptides fall into three families: the insulin-like peptides^{35, 36}. FMRFamide (Phe-Met-Arg-Phe-NH2)-related peptides or FaRPs, which are referred to as FLPs in C. elegans³⁷, and neuropeptide-like proteins or NLPs^{38, 39}. Several conserved neuropeptides are known to play a role in C. elegans associative learning, including homologs for insulin (INS-1) and insulin like peptides (ILPs), oxytocin/vasopressin (NTC-1), myoinhibitory peptide (MIP-1/NLP-38), neuromedin U (CAPA-1/NLP-44), elevenin (SNET-1)^{32, 40–47}.

Neuropeptides function mainly by binding to G protein-coupled receptors (GPCRs). The C. elegans genome encodes an estimated 150 neuropeptidergic GPCRs⁴⁸. Among them, many are orthologs of many evolutionarily ancient neuropeptidergic receptors^{49, 50}, including orthologs of the oxytocin/vasopressin, gonadotropin-releasing hormone (GnRH), thyrotropin-releasing hormone (TRH), neuromedin U (NMU), tachykinin, myoinhibitory peptide (MIP), and neuropeptide Y (NPY) families^{32, 40, 45, 49–51}. Neuropeptidergic receptors offers a rich versatility for the regulation of physiological activities, including metabolism, growth and development, and behaviors in animal organisms, because of diversity of receptors and signal transduction pathways.

There are many neuropeptide-receptor interactions not yet identified in C. elegans, because some peptide-receptor interactions are specific for a pair of proteins, while some peptide-receptor are promiscuous and interact with many partners. Peptide-mediated interactions pose significant challenges for their experimental characterization: These interactions are, in many cases weak, transient, and considerably influenced by their context, resulting in often noisy experiments. Many widely used structure determination methods, such as X-ray crystallography, may not be applicable to a great number of these interactions. To succeed in studying peptide–receptor interactions, we need gain a more effective method to understand how peptide–receptor interactions. Most interactions are determined by the 3D arrangement and the dynamics of the interacting proteins. Recent successful efforts have been made to predict protein structure based on protein sequence, using deep neural network (artificial intelligence, AI) modeling methods, such as OmegaFold, ESMFold and AlphaFold. The methods enable accurate modeling not only of protein monomer structures but also protein complexes^52–58. Therefore, using deep learning methods to predict neuropeptide-receptor interactions may provide a new method for identifying neuropeptide-receptor interactions.

Although, intensive studies have made aiming to reveal molecular and circuital mechanisms underlying short-term SCL. However, pertinent questions remain unclear: what encodes the information of SCL? where is the memory stored?

Here, we used a reverse genetic screen, genetic manipulation, predictions of neuropeptide/receptor interactions by AlphaFold2, quantitative behavior assay, in vivo Ca²⁺ imaging, and neuronal manipulation to study the circuital mechanism underlying short-term SCL in C. elegans. We find that ASER sensation of NaCl is remolded, caused by top-down regulation from interneurons RIC and functionally interaction between ASER and RIS, in and after salt conditioning. ASER sensation remolding or sensory plasticity is required for and tightly correlated with short-term SCL. ASER sensory plasticity encodes the information of short-term salt chemotaxis learning. This finding suggests a new mechanism for learning and memory.

Neuropeptidergic FLP-14/FRPR-10 signaling is required for salt chemotaxis learning

We used a modified test paradigm of salt chemotaxis learning developed by Wicks et al in this study^{59, 60}. For mock and food-deprivation/NaCl association conditioning, about 200 animals of C. elegans were transferred into and immersed in food-uncontained CTX solution without or with 100 mM NaCl for 20 min, respectively. Then, the animals were transferred onto a test agar plate to assay their chemotaxis to 20 mM NaCl (Fig. 1a). The mock-conditioned (treated by mock conditioning) wild-type (WT) N2 animal displayed strong chemotaxis to 20 mM NaCl (salt chemotaxis) indicted by chemotaxis index (CI = (N_B - N_A) / (N_B + N_A)). Here, N_A and N_B are the numbers of animals on the test regions without and with 20 mM NaCl, respectively. While the salt-conditioned (treated by food-deprivation/100 mM NaCl association conditioning) WT N2 animal showed salt avoidance, that is animal acquired salt chemotaxis learning (SCL) (Fig. 1b). We quantified SCL by learning index (LI) that was defined as: CI_mock – CI_conditioned. Here, CI_mock and CI_conditioned are the chemotaxis index of mock-conditioned and salt-conditioned animals, respectively. SCL is a short-term L&M, as it lasts for around 30 min²⁰. We employed the loss-of-function (lof) mutant egl-3(e1370) to primarily examine whether neuropeptides engage in SCL. The gene egl-3 encodes a proprotein convertase, EGL-3, required for the synthesis and processing of neuropeptide precursors⁶¹. The egl-3 mutant displayed the WT CI but reduced LI (Fig. 1c and Supplementary Fig. 1a), indicating that neuropeptides are involved in SCL. We next used 19 flp and 13 nlp mutant animals to screen candidate neuropeptides engaging in SCL. Among the screened animals, the flp-14(gk1055) showed severest defect in SCL, while all the mock-conditioned animals displayed WT salt chemotaxis (Fig. 1d and Supplementary Fig. 1b-e). Therefore, we focused on flp-14 in this study.

The flp-14 gene encodes a propeptide, which is processed to form four same peptide FLP-14 (Supplementary Fig. 1f). The apparent learning deficiency observed in flp-14(gk1055) animal could arise from motor deficit, general defects in salt sensation and chemotaxis, and osmotic sensation. However, flp-14(gk1055) animal’s locomotion speed and chemotaxis to NaCl of varied concentrations did not differ obviously with those of the WT N2 (Supplementary Fig. 1g, h). Moreover, the paring of 200 mM glycerol (osmotic pressure is equal to that of 100 mM NaCl) with food-deprivation did not affect salt chemotaxis in the flp-14(gk1055), contrarily to what did by the paring of NaCl (100 mM) with food deprivation (Supplementary Fig. 2a, b). These results support the behavioral phenotype observed in above experiment is SCL defect.

We next identified candidate receptor of FLP-14. Known physical interactors of FLP-14 include FRPR-19, NPR-1, NPR-4, NPR-6, and NPR-11^{38, 62, 63}. We thus used mutant animals of these receptors to examine their SCL phenotype. Among animals tested, only npr-1(ad609) animal displayed learning defect that could not be restored to the WT by its genetic rescuing expression driven by npr-1 promoter, while all animals showed WT salt chemotaxis (Fig. 1e and Supplementary Fig. 2c). This suggests that an unknown receptor or receptors mediate(s) FLP-14 action in SCL. For labor-saving, we employed AlphaFold 2 (AF2) to anticipate the candidate receptor(s) of FLP-14, by predicting interaction between FLP-14 and its candidate receptors. AF2 is powerful to accurately predict proteins’ 3D structures from their amino acid sequences⁵⁷. As 3D structure of a protein is closely related to its function and interactions with other molecules, AF2 is useful in predicting ligand-receptor interaction^{53, 64–66}. Among the predicted receptor candidates, NPR-1, NPR-2, NPR-5, NPR-7, NPR-10, NPR-17, NPR-24, NPR-39, NPR-41, NPR-42, FRPR-1, FRPR-2, FRPR-4, FRPR-6, FRPR-7, FRPR-9, FRPR-10, FRPR-14, FRPR-15 and FRPR-17 have high scores (≥ 4) of biding with FLP-14 (Supplementary Table 1 and Supplementary Fig. 2d). We assayed the SCL phenotypes in the mutant or knocked-down animals of these receptors. The frpr-10(ok1504) and npr-24(ok3192) showed severe defect in SCL but maintained the WT chemotaxis to NaCl. The SCL defect in frpr-10(ok1504), but not in npr-24(ok3192), was restored to the WT by genetic rescue. RNAi knockdown of frpr-10 but not npr-24 phenocopied genetic mutation (Fig. 1f and Supplementary Fig. 2e-h). AF2 modeling predicts that FLP-14 binds to the FRPR-10 putative-binding pocket in a tight conformation (Fig. 1g). The locomotion speed and salt chemotaxis in the frpr-10(ok1504) and genetically rescued animals did not differ to those in the WT N2 animal (Supplementary Fig. 2i, j). These indicate that FRPR-10 receptor likely functions as a physiological receptor for FLP-14 in SCL.

FRPR-10 is a predicted G protein-coupled receptor⁶⁷. We used a heterogeneous expression system to test the functional interaction between FRPR-10 and its ligand FLP-14, as previously reported^{68, 69}. Briefly, we injected 50 ng of FRPR-10 sense cRNA (complementary RNA) and 0.5 ng of GIRK (G protein-activated inwardly rectifying K⁺)1 and GIRK4 sense cRNA into Xenopus laevis oocytes and recorded whole-cell currents. Application of artificial peptide of FLP-14 generated a robust potassium current in FRPR-10/GIRK1/GIRK4 cells in a dose-dependent manner with a half maximal effective concentration (EC₅₀) of 3.7 µM, but no apparent current in GIRK1/GIRK4 cells (Fig. 1h, i). The apparent EC₅₀ seems high, possibly because of a low purity of the peptide product.

To further determine whether FLP-14 peptide functions through FRPR-10 in SCL in vivo, we constructed flp-14 and frpr-10 double knocked-out (DKO) animal and examined its SCL phenotype. The DKO animal displayed a defect in SCL but not in salt chemotaxis, similarly to the flp-14(gk1055) and frpr-10(ok1504) animals. The defect could not be rescued by single reconstitution of flp-14 or frpr-10 and was restored to the WT by the simultaneous rescue of both genes (Fig. 1j and Supplementary Fig. 2k). In conclusion, above tests support that FRPR-10 is the receptor of FLP-14 and FLP-14/FRPR-10 signaling is required for SCL.

ASER-RIS circuit engages in SCL

Behavioral responses to sensory stimulations depend on neural information encoding, decoding, and integration in neuronal circuits. Circuital mechanism studies are essential for neuroscience. We set about our circuital studies. FLP-14 is expressed in main and secondary salt-sensitive gustatory ASE (Amphid Single Cilium E) and ASG (Amphid Single Cilium G) and interneuron RIS (Ring Interneuron S), identified by the expression pattern of GFP driven by a 2kb flp-14 promoter (Supplementary Fig. 3a-e). The reconstitution of flp-14 in its expression cells or ASER alone, but not in ASEL, ASG, or RIS, fully recovered the WT SCL behavior (Fig. 2a and Supplementary Fig. 3f), indicating flp-14 functions in ASER gustatory neuron. We next used neuron-specific flp-14 knockdown by RNA interference (RNAi) and neurotransmission elimination by TeTx to validate flp-14 functional site. TeTx, a light chain of tetanus toxin, is a specific protease of synaptobrevin⁷⁰, that has been successfully used to inhibit chemical synaptic transmission in tested neurons (neuron::TeTx in short) in C. elegans^71–77. As expected, ASER-specific flp-14 knockdown reduced SCL ability in the salt-conditioned transgenic animal but had no obvious impact on salt chemotaxis in the mock-conditioned animal. However, the ASER::TeTx manipulation not only suppressed SCL but also salt chemotaxis (Fig. 2b, c, and Supplementary Fig. 3g, h).

FRPR-10 is expressed in single interneuron RIS, identified by the expression pattern of GFP driven by frpr-10 promoters with varied length of 1 kb, 2 kb, and 4 kb (Supplementary Fig. 4a-d). The frpr-10 reconstitution in RIS driven by promoters of frpr-10 (2 kb) and flp-11 (2 kb) fully rescued SCL defect (Fig. 2d and Supplementary Fig. 4e). Moreover, the RIS::TeTx manipulation and RIS-specific frpr-10 knockdown significantly reduced the learning ability (Fig. 2e, f, and Supplementary Fig. 4f, g). These support that FLP-14/FRPR-10 signaling mediates neurotransmission from ASER to RIS and engages in SCL.

What is the effect of FLP-14/FRPR-10 signaling on RIS activity and underlying mechanism? To answer this issue, we measured RIS Ca²⁺ signals in response to 50 mM NaCl stimulation after NaCl/food-deprivation association conditioning. No obvious change in Ca²⁺-responses was observed (Supplementary Fig. 5a-c). FRPR-10 receptor is a predicted G protein-coupled receptor⁶⁷. It is possible that FRPR-10 signaling does not affect RIS excitability indicated by Ca²⁺ signals but regulates neurotransmitter and/or neuropeptide release by activation of G proteins. G proteins are usually expressed extensively in various tissues or cells. We thus used RIS specific RNAi knockdown of pan-neuronally expressed genes encoding α-subunit of heterotrimeric G proteins, gsa-1 (G_s), egl-30 (G_q), and goa-1 (G_i)⁷⁸, to preliminarily identify the potential signaling pathway. The RIS-specific RNAi knockdown of gsa-1 and its effect gene acy-1 (encoding adenylate cyclase-1, ACY-1), but not egl-30 or goa-1, decreased learning index or suppressed SCL (Supplementary Fig. 5d, e). This suggests that FRPR-10 functions through Gs-ACY-1 signaling pathway (Supplementary Fig. 5f).

Homologous comparison indicated that FRPR-10 is highly homologous with the neurotensin receptor NTSR1 in mammals (Supplementary Fig. 6a). NTSR-1 functions in regulating metabolism, multiple physiological activities, and behaviors including sleep, working memory, fear learning, and memory consolidation^79–83. For examining the functional conservativeness of FRPR-10, we used ntsr-1 cDNAs to conduct genetic rescue test. Interestingly, expression of both rat and human ntsr-1 in RIS driven by frpr-10p (2 kb) partially and significantly rescued SCL phenotype on the background of frpr-10 mutant (Fig. 2g and Supplementary Fig. 6b). This suggests that FRPR-10/NTSR-1 is high conservative through evolution.

FLP-14/FRPR-10 signal pathway in RIS mediates ASER sensatory plasticity in SCL process

ASEs are the main salt sensory neurons functioning in chemotaxis to water soluble attractants including Na⁺, Cl⁻, cAMP, biotin, and lysine. Sensory information is the base of learning and memory. We thus examined ASE sensory responses to NaCl after mock and salt conditioning, using Ca²⁺ fluorescence imaging. ASER Ca²⁺ signals in response to stimulation of 20 mM and 50 mM NaCl in the mock-conditioned WT N2 animal displayed a robust hyperpolarization (inhibitory) ON response and an even stronger depolarization (excitatory) OFF-response. Interestingly, the ASER Ca²⁺ signals of both ON- and OFF-responses in the salt-conditioned animal were nearly disappeared. The intensities of ON- and OFF-responses to 50 mM NaCl were stronger than those to 20 mM NaCl (Fig. 3a-c and Supplementary Fig. 7a-c, e). Thus, we used 50 mM NaCl to be stimulation in this study. This result indicates that ASER salt sensation is remolded in SCL process. We then introduced sensory plasticity index to indicate the ASER sensation change. The sensory plasticity index of ON- and OFF-responses to 50 mM NaCl stimulation in the salt-conditioned WT N2 animal was similar and around 0.8 and bigger than those of responses to 20 mM NaCl stimulation (Fig. 3d and Supplementary Fig. 7d). Does ASEL activity change after the salt conditioning like ASER? ASEL only displayed a robust ON-response but no obvious OFF-response to 50 mM NaCl stimulation in the mock- and salt-conditioned WT N2 animal. Moreover, ASEL Ca²⁺-response in the salt-conditioned animal almost copied that in the mock-conditioned animal. (Fig. 3e-g). These indicate that ASEL salt sensation is not modulated in the SCL behavior.

Given that FRPR-10 signaling in RIS is essential for SCL, a logical issue is that this signaling very likely mediates sensory plasticity in ASER. Expectedly, ASER sensory plasticity in the salt-conditioned frpr-10(ok1504) animal was dramatically reduced and fully restored to the WT by frpr-10 reconstitution in RIS (Fig. 3h-j and Supplementary Fig. 7f). Moreover, ASER sensory plasticity in the salt-conditioned flp-14(gk1055) and ASER-specific flp-14-rescued animals almost phenocopied that in the frpr-10 mutant and RIS-specific frpr-10 rescued animals (Fig. 3k-m and Supplementary Fig. 7g)

Basing on above results, we conclude that FLP-14/FRPR-10 signaling is not only required for SCL but also for ASER plasticity and that sensory plasticity may be a mechanism of the short-term learning and memory.

Neuropeptide PDF-2 released from RIS up-regulates SCL and ASER sensory plasticity

Given that FRPR-10 signaling in RIS are required for both SCL and ASER plasticity. RIS should regulate ASER sensory plasticity in the process of SCL. What is the mechanism for RIS regulating SCL and ASER sensory plasticity? Single interneuron RIS is GABAergic and neuropeptidergic. It is presynaptic to chemically or electrically connects with locomotion-regulating neurons, such as AIB (Anterior Interneuron B) (electrically), AVE (Anterior Ventral Process E), RIM (Ring Interneuron M), SMD (Sublateral Motor Neuron D), and motor neurons DB (Dorsal B-type Motor Neuron). However, there is no known synaptic connection between RIS and ASE^{84, 85}. Thus, RIS may modulate ASER activity by neurohumoral regulation. For analyzing RIS function in SCL, we used RIS-specific RNAi knockdown of unc-25 (encodes a glutamic acid decarboxylase the essential for GABA biosynthesis), egl-3 (encodes a proprotein convertase), inx-6, inx-7, and unc-9 (encode C. elegans gap-junction proteins, expressed in RIS)^{61, 86–88}, to test SCL behavior in the transgenic animals. All transgenic animals showed reduced learning index or decreased SCL after salt conditioning and the WT salt chemotaxis after mock conditioning (Supplementary Fig. 8a, b). This indicates that chemical and electrical synapses connections and neuropeptide(s) from RIS are required for SCL. For detecting RIS regulatory role in ASER sensory plasticity, we used lof mutant animals of unc-25, egl-3, and unc-9 to assay ASER Ca²⁺-responses to 50 mM NaCl stimulation after mock and salt conditioning and to analyze ASER sensory plasticity. As shown in Supplementary Fig. 8c-n, the egl-3(gk238) but not unc-25(ok1901) or unc-9(fc16), displayed significant changes in ON- and OFF-responses to the salt stimulation and defect of sensory plasticity after salt conditioning but not mock conditioning, in comparison with those of the WT N2. Above results suggest that synaptic neurotransmission from RIS is required only for SCL but not for ASER sensory plasticity, while RIS neurohumoral regulation is essential for both SCL and ASER sensory plasticity.

RIS expresses neuropeptide-encoding genes pigment dispersing factor 2, pdf-2, also named nlp-37⁸⁹. The nlp-37(tm4393) animal displayed a significant defect of salt chemotaxis learning after salt conditioning and exhibited the WT chemotaxis to NaCl after mock conditioning (Supplementary Fig. 1d, e), suggesting PDF-2 or NLP-37 may mediate SCL. PDF-2 receptor PDFR-1 is expressed in muscle cells and multiple neurons including ASER⁹⁰. We thus used genetic analyses of pdf-2 and pdfr-1 to detect the roles of these two genes in SCL and ASER sensory plasticity. As expected, lof mutation of pdf-2 reduced SCL index that was restored to the WT by genetic rescue driven by both pdf-2p and RIS-specific promoters (Fig. 4a and Supplementary Fig. 9a). Although lof mutation of pdfr-1 decreased SCL ability that rescued to the WT by genetic reconstitution driven pdfr-1p. However, the pdfr-1 reconstitution in ASER alone could partially rescue the behavioral defect (Fig. 4b). pdfr-1 expressed extensively in muscle cells and neurons, including sensory neurons, interneurons, and motor neurons. pdfr-1 mutant animals showed defect in locomotion⁹¹. Thus, we examined the SCL, chemotaxis, and locomotion phenotypes in the ASER-specific pdfr-1 RNAi knocked-down animal. The pdfr-1 knockdown impaired SCL like RIS-specific knockdown of pdf-2 but did not affect chemotaxis and locomotion speed (Fig. 4c and Supplementary Fig. 9b-d). These results suggest that PDF-2/PDFR-1 signaling act on ASER to regulate SCL behavior.

PDF-2/PDFR-1 signaling should regulate or control ASER sensory plasticity. We used animals of pdf-2 and pdfr-1 mutation and transgenic rescue to detect ASER Ca²⁺-responses to 50 mM NaCl stimulation and analyze sensory plasticity after mock- and salt-conditioning. Expectedly, ASER in the salt-conditioned pdf-2 or pdfr-1 animals showed robust ON- and OFF-responses to the salt stimulation which were almost eliminated by genetic reconstitution of pdf-2 in RIS or pdfr-1 in ASER. ASER sensory plasticity in the salt-conditioned mutant animals were almost disappeared and was restored to the WT by the genetic rescue of pdf-2 in RIS or pdfr-1 in ASER (Fig. 4d-i and Supplementary Fig. 10a, b). In summary, PDF-2/PDFR-1 signaling pathway mediates neurohumoral regulation of ASER by RIS in top-down regulation of SCL and ASER sensory plasticity.

RIC inhibits and excites ASER and AIY in the regulation of salt chemotaxis learning by tyraminergic TYRA-2 and octopaminergic OCTR-1 signaling

Food deprivation is a conditioning stimulation for SCL. The mechanism of sensory information integration and translation into behavior needs study. Paired RIM and mainly RIC interneurons service as a center for integrating sensory information of stress and starvation or food deprivation^{74, 92–98}. RIM and RIC release TA and OA, respectively. Octopamine released from RIC leads to activation of the transcription factor CREB required for long-term learning and memory^{99, 100}. Mammal orthologue of TA and OA, adrenaline and noradrenaline, engage in long-term associative learning and memory in mammals^101–103. TA is synthesized from tyrosine catalyzed by tyrosine decarboxylase TDC-1. OA is transformed from TA catalyzed by tyramine-beta-hydroxylase TBH-1. The genes tdc-1 and tbh-1 encode TDC-1 and TBH-1 respectively. tdc-1 is expressed in RIM and RIC, while tbh-1 is mainly expressed in RIC¹⁰⁴. Thus, we used tdc-1 and tbh-1 mutant animals to test their SCL phenotype. Expectedly, SCL ability reduced in the tbh-1 and tdc-1 mutant animals without change in chemotaxis to NaCl after mock conditioning. The behavioral defect in the mutant animals were eliminated by genetic reconstitution of each gene driven by each promoter. However, the tdc-1 reconstitution in RIC but not in RIM eliminated the behavioral defect caused by the tdc-1 mutation (Fig. 5a and Supplementary Fig. 11a). This suggests RIC but not RIM engages in SCL. We next used pharmacological test in the tdc-1; tbh-1 double knockout (DKO) animal to identify the function of TA and OA or RIC and RIM neurons in SCL. Administration of TA, OA, or both had no obvious impact on SCL and salt chemotaxis in the WT N2 animal. Puzzlingly, neither the treatment of OA nor that of TA, but the application of both OA and TA restored the WT SCL in the DKO animal (Fig. 5b and Supplementary Fig. 11b). Possibly, RIC releases both TA and OA and has more than one function sites. We then used neurotransmission inhibition by RIM::TeTx, RIC::TeTx, and RIM/RIC::TeTx to validate above results. RIC::TeTx and RIC/RIM::TeTx similarly suppressed SCL, while RIM::TeTx had no obvious impact on the behavior (Fig. 5c and Supplementary Fig. 11c). In summary, these tests indicate that RIC but not RIM engages in SCL, and RIC possibly has multiple function effect cells through both TA and OA signal pathways.

Does RIC activity change after salt/food-deprivation conditioning? To answer this issue, we performed RIC Ca²⁺-imaging with G-CaMP3.0 as an indicator. RIC displayed a minor excitatory ON-response of Ca²⁺ transients in the mock-conditioned WT N2 animals. After salt/food-deprivation association conditioning, the RIC Ca²⁺-response almost disappeared and significantly differed with that in the mock-conditioned animal (Fig. 5d and Supplementary Fig. 11d). This suggests that RIC activity in response to the NaCl stimulation is inhibited after the salt conditioning and support RIC engages in SCL.

What are TA and OA receptors functioning in SCL? Among mutant animals of known TA and OA receptors, TYRA-2, TYRA-3, SER-2, LGC-55, OCTR-1, SER-3, and SER-6, the tyra-2(tm1846), tyra-2(tm1815), and octr-1(ok371) displayed obvious behavioral defect, while all animals showed normal salt chemotaxis. SCL defect caused by tyra-2 and octr-1 lof mutation was removed by the genetic reconstitution of each gene driven by its promoter (Fig. 5e, f and Supplementary Fig. 11e-h). Tyraminergic receptors TYRA-2 is expressed in cells including ASER and AIY that functions in chemotaxis and locomotion, while OCTR-1 is expressed in cells including AIY (https://wormbase.org/). It is possible that TYRA-2 and OCTR-1 signaling function in ASER and AIY to regulate SCL. We thus used genetic reconstitution of these two genes to identify their functional sites. Our test result showed that among multiple neuron(s)-specific constitution of tyra-2 or octr-1, rescue expression of tyra-2 in ASER (driven by 3.2 kb gcy-5p) and octr-1 in AIY (directed by 4.0 kb T19C4.5p) fully restored the WT SCL in the transgenic animals (Fig. 5e, f and Supplementary Fig. 11i, j). We then used neurotransmission inhibition by AIY::TeTx to validate whether AIY functions in SCL and found AIY::TeTx impaired SCL (Supplementary Fig. 11l, m). These results support that tyraminergic TYRA-2 and octopaminergic OCTR-1 signaling act in ASER and AIY to generate SCL behavior. To assay AIY Ca²⁺ responses to the NaCl stimulation in mock- and salt-conditioned animals, we performed Ca²⁺-imaging with G-CaMP3.0 as an indicator. AIY displayed only a minor inhibitory and excitatory ON-response of Ca²⁺ transients in the mock- and salt-conditioned WT N2 animals, respectively (Supplementary Fig. 11n-p). Combine with the suppressed RIC Ca²⁺-response to NaCl stimulation in salt-conditioned animal, this suggests that RIC may inhibit AIY activity by OA/OCTR-1 signaling, RIC suppression by salt-conditioning removes the inhibition of AIY by RIC, and thus alter chemotaxis locomotion.

ASER and RIS form a negative feedback circuit essential for SCL, in which ASER possibly activates RIS to release neuropeptide PDF-2 through FLP-14/FRPR-10 signaling and RIS inhibits ASER activity via PDF-2/ PDFR-1 signaling (Fig. 2, 3). RIC functions in SCL through TYRA-2 and OCTR-1 acting in ASER and AIY, respectively (above result). Thus, RIC, ASER, and RIS may form a functional circuit to regulate SCL behavior. For answering this issue, we analyzed the functions of TYRA-2 and FRPR-10 signaling pathways acting in ASER in SCL, using genetic analyses of tyra-2 and frpr-10. Double knockout of tyra-2 and frpr-10 impaired SCL like that of single knockout of tyra-2 or frpr-10. SCL phenotype was restored to the WT only by genetic reconstitution of both tyra-2 and frpr-10, not by that of single gene of tyra-2 or frpr-10 (Fig. 5g and Supplementary Fig. 11k). This supports that RIC, ASER, and RIS form a circuit to modulate the short-term SCL behavior.

ASER sensory plasticity encodes the information of the short-term salt chemotaxis learning

TYRA-2 signaling functions in SCL, it should act on ASER sensory plasticity. We thus employed genetic analyses of tdc-1 and tyra-2 to answer this issue. ASER in the mock-conditioned tdc-1(n3419), tyra-2(tm1846), and RIC::tdc-1 and ASER::tyra-2 gene-rescue animals displayed the WT ON- and OFF-responses to the NaCl stimulation. While ASER in the salt-conditioned tdc-1 and tyra-2 mutant animals show obvious ON- and OFF-responses to the NaCl stimulation and significantly reduced sensory plasticity which are restored to the WT phenotypes by the gene-rescue of tdc-1 in RIC and tyra-2 in ASER, respectively (Fig. 6a-f and Supplementary Fig. 12a, b). Moreover, the effect of RIC::TeTx genetic manipulation on ASER Ca²⁺-responses to the NaCl stimulation and sensory plasticity fully phenocopied that of tdc-1 lof mutation (Fig. 6g-i and Supplementary Fig. 12c). These results indicate that RIC controls ASER sensory plasticity under salt-conditioning through TA/TYRA-2 signaling.

Basing on SCL phenotypes and ASER activity changes, we can deduce that inhibition of ASER’s sensation of NaCl or ASER sensory plasticity is related with SCL. Thus, we analyzed correlation between ASER’s sensory plasticity and animal’s learning indexes in the tested animals, including the WT N2, the mutant and transgenic animals of flp-14, ASER::flp-14, frpr-10, RIS::frpr-10, pdf-2, RIS::pdf-2, tdc-1, RIC::tdc-1, tyra-2, ASER::tyra-2, and RIC::TeTx. ASER’s sensory plasticity are well fitted with learning indexes by linear regression (R squared = 0.8550 and p < 0.0001) (Fig. 6j). This support that the ASER’s sensory plasticity is positively and directly correlates with the short-term salt chemotaxis learning. Very interestingly, ASER’s sensory plasticity and the short-term SCL are all-or-none. ASER’s sensory plasticity is possibly a cellular mechanism for short-term salt chemotaxis learning.

Our results support that the signaling of FLP-14/FRPR-10 (mediating neurotransmission form ASER to RIS, Fig. 3h-m), PDF-2/PDFR-1 (mediating neurotransmission form RIS to ASER, Fig. 4d-i), and TA/TYRA-2 (mediating neurotransmission from RIC to ASER, Fig. 6a-i), regulate ASER sensation of NaCl or required for ASER sensory plasticity in SCL. Is there difference in the effect of these signaling on ASER sensory plasticity? Shown in Fig. 6k, the analysis indicates that the ASER sensory plasticity in salt-conditioned mutant animals of flp-14, frpr-10, pdf-2, pdfr-1, tdc-1, and tyra-2, and the transgenic RIC::TeTx animal displayed no obvious difference. This supports that all signaling pathways are required for ASER sensory plasticity and that RIC, ASER, and RIS form a circuit to generate the short-term SCL behavior.

Hitherto, this study identifies that sensory neuron ASER and interneurons RIC and RIS paly pivot roles in the short-term SCL. What is the role of each neuron in the processes of learning and memory? For answering this issue, we used chemogenetics to acutely inhibit these three type neurons at learning or memory-recalling stage, by using cell-specific expression of HisCl1 channels and a treatment of 10 mM histamine. HisCl1 is a histamine-gated chloride channel subunit from Drosophila that is effective to tempo-spatially inhibit neurons when activated by exogenous histamine^{72–75, 105, 106}. Our results showed that the chemogenetic inhibition of ASER in both learning and memory recalling stages decreased the learning index of SCL, while the inhibition of RIC or RIS in the learning stage but not in the memory recalling stage reduced SCL ability (Fig. 6l). Whereas the chemogenetic inhibition of all three type neurons in both learning and memory-recalling stages did not affect chemotaxis to NaCl in the mock-conditioned animals (Supplementary Fig. 12d). This result indicates that ASER engages in both learning and memory recalling and both RIC and RIS interneurons are involved in only learning but not memory recalling.

The short-term SCL needs about 10 min for the induction of the behavioral plasticity and lasts for about 30 min²⁰. Our results suggest that ASER sensory plasticity may be a major cause for the short-term SCL behavior. Thus, the time course of ASER sensory plasticity should be like that of the behavior. Indeed, assayed by Ca²⁺ imaging, the salt-conditioning of about 15 min caused significant ASER sensory plasticity. The loss-of-function of TDC-1 and FLP-14 fully suppressed the induction of ASER sensory plasticity (Fig. 7a-e and Supplementary Fig. 13a-c). Moreover, ASER sensory plasticity lasted for about 30 min (Fig. 7f-h and Supplementary Fig. 13d). These results support that the time course of ASER sensory plasticity is the same as that of the short-term SCL. In conclusion, all our test results support that ASER sensory plasticity encodes the information of the short-term SCL.

Here, we identify a circuital mechanism of short-term SCL. The main salt sensory cell ASER and interneurons RIC, RIS, and AIY play pivotal roles in the SCL. We further reveal the underlying mechanism of SCL. In the circuit, RIC, ASER, and RIS form a circuit to remodel ASER sensation of NaCl or cause ASER sensory plasticity in salt-conditioning process. RIC integrate signals of food availability and NaCl. It releases OA and TA and may activate AIY and augment ASER sensory response to NaCl by acting on receptors OCTR-1 on AIY and TYRA-2 on ASER, respectively. RIC Ca²⁺-response to NaCl stimulation is inhibited in the salt-conditioned animal. Thus, RIC activating AIY and augmenting ASER are reduced in the salt-conditioned animal (Fig. 7i). However, the mechanism for RIC integrating sensory signals and remolding response to the NaCl stimulation in the salt-conditioned animal needs further study.

The inhibition of ASER salt sensation or ASER sensory plasticity generated by salt conditioning encodes the information of short-term salt chemotaxis learning

ASEL and ASER are functionally specialized in NaCl sensation and NaCl chemotaxis^107–109. ASEL is an ON-cell, stimulated by increases in NaCl concentration, whereas ASER is an OFF-cell, stimulated by decreases in NaCl concentration¹⁰⁹. ASEL and ASER positively and negatively regulates the chemotaxis to NaCl by augmenting forward locomotion and causing turns, respectively¹⁰⁹. ASEL and ASER differently regulate salt chemotaxis learning with different signaling pathways^{40, 110–113}. Here we find that ASER is a both ON- and OFF-cell of NaCl sensation, with a robust ON-response and an even stronger OFF-response at least under the mock (no-NaCl/food-deprivation association) conditioning and our test conditions. Our explanation is that Suzuki et al. focused on the OFF-response of ASER NaCl sensation. They also recorded a small stepwise and sustained hyperpolarization Ca²⁺ signal change in response to a switch of NaCl concentration from 40 mM to 80 mM. ASER sensory responses to NaCl in the mock- and salt-conditioned animals observed in this study slightly differs from those of previous reports^{45, 109, 111, 114}. This difference possibly caused by different conditioning and measuring: the association of 100 mM NaCl with food deprivation for 20 min for salt-conditioning and the switch of 0–50 (30s) -0 mM NaCl for NaCl stimulation in this study; the association of 20 mM NaCl with food deprivation for 10 min and the switch of 0–20 mM NaCl in the study by Odo et al.¹¹¹; similar conditioning and a 0-50mM NaCl upshift in the work by Watteyne et al. ⁴⁵; 20 mM NaCl with food deprivation for at least 180 min and 0-20-0 mM NaCl switch in the work by Lim et al. ¹¹⁴.

Of note, ON- and OFF-responses to NaCl stepwise increase and decrease n in ASER in the mock conditioned animal are hyperpolarization and depolarization. For the ASER excitability and activities, especially vesicular release, ON- and OFF-responses are inhibitory and excitatory, and thus the sensation inhibition or sensory plasticity of these responses are excitatory and inhibitory, respectively. Moreover, ON- and OFF-responses have different dynamics: sustained (or not-decayed) and transient, respectively. The difference of two phases of ASER NaCl sensory responses may endow different functions of two phases in salt chemotaxis: inhibitory ON-response, inhibiting direction switch; excitatory OFF-response, augmenting turn or direction switch. Possibly, ON-response is more important than OFF-response in C. elegans NaCl chemotaxis. Importantly, our study finds that ASER NaCl-sensation in the salt-conditioned animals is almost fully suppressed compared with those in the mock-conditioned animals, that is salt conditioning causes ASER sensory plasticity. The facts that ASER sensory plasticity is positively and linearly corelated with SCL behavior, it lasts for about 30 min, and chemogenetic inhibiting ASER at both learning and memory stages significantly impairs SCL (Figs. 6 and 7), support that ASER sensory plasticity is required for SCL. It is reasonable to conclude that the sensory plasticity may encode the information of short-term SCL.

ASER plays a major role in NaCl chemotaxis by both pirouette and weathervane behavioral strategies with ASEL minor contribution. Laser ablation of ASER but not ASEL reduces chemotaxis to NaCl, while salt conditioning reverses NaCl attraction to NaCl aversion^{115, 116}. This support signal from ASER is essential for SCL. The loss of function of neuropeptide FLP-14 and its receptor FRPR-10 equally suppresses SCL (Figs. 1 and 2) and reduces ASER sensory plasticity (Figs. 3 and 7), indicating FLP-14/FRPR-10 signaling is essential for SCL and ASER sensory plasticity. This study discovers that ASER sensory plasticity is essential for short-term salt chemotaxis and encodes the information of the behavior. ASE neurons send major synaptic outputs to three interneurons, AIA, AIB, and AIY. Of these, laser ablation of AIZ caused severe defect in salt chemotaxis¹¹⁵. The roles and activity changes of these neuron in short-term SCL and underlying mechanisms need further study.

RIC regulates AIY activity and generates ASER sensory plasticity through OA and TA signaling to generate SCL

RIC seems to be interneuron integrating information of external and internal stress and food-supply especially food-deprivation, and internal stress and functions in avoidance behavior, feeding, fat metabolism, behavioral plasticity^{72, 74, 92, 93, 95, 97, 117}. It may function as a counterpart of mammalian adrenal gland. RIC activity is inhibited by food-sensing neuron ADF in a negative correlation manner with food-supply and may synthesize and release both TA and OA⁷⁴, TA and OA are invertebrate counterparts of adrenaline and noradrenaline. Adrenergic transmitters function as modulators for metabolism, cellular immune response, the modulation of sensory input, and multiple behaviors including the fight or flight response and learning and memory¹⁰². In this study, we identify that RIC release both TA and OA to regulate AIY interneuron and ASER sensory neuron by acting on OCTR-1 and TYRA-2 receptors, respectively.

AIY inhibits direction-switch by forming an inhibitory circuit with the downstream interneuron AIZ that promotes direction switch and promotes locomotion speed by forming an excitatory circuit with RIB¹¹⁸. RIC Ca²⁺-response to NaCl stimulation in the salt-conditioned animal was fully suppressed. tdc-1 in RIC, octr-1 in AIY, and both TA and OA are required for normal SCL (Fig. 5). These indicate that OA from RIC may function to inhibit AIY through OCTR-1 signaling and AIY inhibition by RIC is reduced when the salt-conditioned animal reencountered NaCl. OCTR-1 is an ortholog of human ADRA2A (adrenoceptor alpha 2A) implicated in several diseases, including artery disease; attention deficit hyperactivity disorder; and type 2 diabetes mellitus⁶⁷. OA/OCTR-1 signaling acts on AIY to augment avoidance of high osmolarity⁷⁵. Through the OCTR-1 receptor, OA activates Gα_o signaling resulting in GIRK channel activation, presumably through the release of G_βγ subunits, thus inhibiting the initial 1-octanol dependent depolarization¹¹⁹. Interestingly, octopamine acts on ASH via OCTR-1 to decrease aversion of bacterially produced volatiles such as octanol¹²⁰. AIY may employs different signaling in avoidance behaviors in responses to varied cues. For example, AIY uses SER-6 signaling in learned bacterial aversion⁹⁵.

TYRA-2 engages in SCL, as demonstrated in both Moro et al²³ and this study. Noteworthily, this study revealed that TA/TYRA-2 signaling modulates ASER sensation of NaCl or causes sensory plasticity in the short-term salt-conditioning. Yet, the underlying mechanism remains elusive and needs further study. TYRA-2 is a Gα_i/o-coupled tyramine receptor¹²¹. Neurohumoral TA from RIM potentiates ASH sensitivity to osmotic stimuli through TYRA-2 signaling¹²². TA/TYRA-2 signaling inhibits RIM activity and mediates feeding suppression by repellent 1-octanol¹²³. TYRA-2 is required for avoidance responses to osas#9, an ascaroside pheromone, possibly by acting as a receptor for osas#9¹²⁴, however sensation modulation by TYRA-2 cannot be excluded.

Here, we identify that TA/TYRA-2 signaling augments ASER salt sensation. The suppress of TA release in RIC by salt conditioning co-acts with FLP-14/FRPR-10 and PDF-2/PDFR-1 signaling pathways to generate ASER sensory plasticity in the salt conditioning process.

Neuropeptidergic FLP-14/FRPR-10 and PDF-2/PDFR-1 signaling pathways in ASER-RIS feedback circuit are required for short-term salt chemotaxis learning

Signaling Pathways of TA/TYRA-2, FLP-14/FRPR-10, and PDF-2/PDFR-1 are equally required for ASER sensory plasticity in and after salt-conditioning (Fig. 6k and 7a-e). ASER and RIS form a feedback circuit. ASER secretes neurohormone FLP-14, a FMRFamide (Phe-Met-Arg-Phe-NH2)-related peptide, to activate RIS through FRPR-10-GSA-1(Gα_s)-ACY-1(adenylyl cyclase) signaling (Supplementary Fig. 5). RIS releases neurohumoral PDF-2 acting on PDFR-1 on ASER to inhibit ASER sensory response to the NaCl stimulation in the salt conditioned animal. As ASER ON- and OFF-responses are inhibitory and excitatory respectively, feedback regulation from RIS to ASER may function differently: augmenting and reducing ASER release of neurotransmitters and modulators in ON and OFF sensory phases. ASER ON- and OFF-responses may function differently in chemotaxis to NaCl and short-term SCL. This issue needs further study.

Neuropeptidergic receptor FRPR-10 exhibits sequence homology to the vertebrate neurotensin receptors-1, NTSR-1 (Supplementary Fig. 6). Moreover, heterologous expression of ntsr-1 cDNA from rat and human driven by frpr-10 promotor in C. elegans functions well in short-term SCL as frpr-10 does (Fig. 2 and Supplementary Fig. 6). These support that FRPR-10 is a homolog of mammalian NTSR-1. Neurotensin/NTSR-1 signaling engages in learning and memory in mammals including human and neurotensin may function as an important neuromodulator^{82, 125–128}. Mounting evidences support that neuropeptides are important to learning and memory in animals^129–136. This study discovers that FLP-14 engages in short-term SCL and its receptor FRPR-10 is conservative through the evolution. Of note, PDF2/PDFR-1 signaling pathway mediates top-down regulation from interneuron RIS to sensory neuron ASER that required for ASER sensory plasticity in salt conditioning and short-term SCL.

A neuropeptide may bind with multiple receptors and a neuropeptidergic receptor may bind with multiple neuropeptides. Traditional experimental screen to identify ligand-receptor pairs are labor and time consuming. In identifying receptor(s) of FLP-14 that mediate(s) ASER action on RIS, we used AlphaFold2, a deep-learning-based structure prediction method, to predict potential receptors by modeling ligand-receptor interaction and structure of receptors. AlphaFold2 is useful for the modeling of protein–protein complexes and predicting protein–protein interaction in proteomes of human and yeast^{53, 55, 137}. Among nineteen FLP-receptor interactions with the number of model match ≥ 4, successfully predicted FRPR-10 functions in the short-term SCL. It does not mean other predicted receptors are surely not physiological receptors, they may engage in other functions. The prediction largely narrows our screening scope.

Single RIS interneuron is a sleep-active neuron, which induces behavioral quiescence and physiological restoration. At sleep onset RIS depolarizes and releases FLP-11 to induce a systemic sleep state^{138, 139}. Additionally, RIS is required for survival of starvation and wounding and functions as a locomotion stop neuron.^{140, 141} Sleep is important for memory consolidation and benefits the retention of memory¹⁴². Insufficient sleep impairs memory acquisition and retention in both vertebrates and invertebrates^143–145. In C. elegans sleep is required for odor memory consolidation and synaptic remodeling, ALA promoting sleep is necessary for long lasting memory, and AIY that functions in odor seeking is required memory consolidation¹⁴⁶. This study finds that interneurons RIS, RIC are required for memory acquisition of the short-term SCL but not for memory recalling (Fig. 6i).

In summary, this study finds that ASER sensory plasticity generated by top-down regulation by RIC and RIS encodes the information of salt chemotaxis learning. Learning and memory, an adaptive behavior plasticity, is based on physiological and structural plasticity of the nervous system. The plasticity of neurotransmission—long-term potential and long-term depression, and structural plasticity—formation and maintenance of synapses, pave neuronal basis for long-term learning and memory^{1, 2, 147}. Short-term depression and facilitation of synaptic transmission by sensory neurons mediate short-term habituation and sensitization². Here, we find that the plasticity of Ca²⁺-responses to NaCl stimulation in sensory neuron caused by top-down regulation is required for the short-term SCL. Since calcium activity reflects any neural activity, including synaptic release, the plasticity of Ca²⁺-responses presents sensory plasticity. This finding reveals a new model for short-term learning and memory.

Maintenance and generation of C. elegans strains

All C. elegans strains were cultured on nematode growth media (NGM) plates at 20°C using E. coli bacteria OP50 as food by standard procedures. The strains were obtained from the CGC (http://www.cbs.umn.edu/CGC/) or the National Bio-Resources Project (NBRP, http://www.shigen.nig.ac.jp/c.elegans/index.jsp). All transgenic animals were generated with standard microinjection techniques¹⁴⁸. The most plasmids were injected at 50 ng/µL together with lin-44p::GFP or vha-6p::GFP as a coinjection marker at 10 ng/µL. The injection pressure was controlled by a DMP-300 digital pneumatic microinjection pump (Micrology Precision Instruments, Ltd, Wuhan, China). The double mutant animals were generated using the standard genetic techniques¹⁴⁹, and confirmed by PCR and sequencing. The strain of frpr-10(ok1504) was backcrossed to the wild-type (WT) N2 five times to remove unlinked mutations. The strain of flp-14(gk1055) was backcrossed to the WT N2 five times to remove sfxn1.2 mutation. All animals used in this study are listed in supplementary table 2.

Molecular biology

Construction of expression plasmids. All expression plasmids used in this study were constructed by Three-Fragment Multisite® gateway system (Invitrogen™, Thermo Fisher Scientific, Waltham, MA, USA). Briefly, three entry clones, A, B and C, comprising three PCR products (promoter, the gene of interest, and sl2::TagRFP-t or 3’ UTR, respectively) were recombined into the pDEST™ R4-R3 Vector II (from Addgene, http://www.addgene.org/vector-database) or custom-modified destination vectors pDEST-R4-R3-unc-54 3’UTR or pBCN44-R4-R3-lin-44p::GFP, using attL-attR (LR) recombination reactions to generate expression clones. pDEST-R4-R3-unc-54 3’UTR destination vector was modified from the pDEST-R4-R3 vector by inserting unc-54 3’ UTR at the position between attR3 and AmpR promoter, with PCR-linearization and In-fusion methods. pBCN44-R4-R3-lin-44p::GFP destination vectors was modified from the pBCN44-R4-R3 vector by inserting lin-44p::GFP serving as a gene marker for transgene into the backbone with restriction enzyme digestion method.

Construction of an entry clone A containing a promoter. An entry clone A (or promoter vector) contributing sequence of slot 1 in the expression plasmid contains sequence of a promoter. We used an In-Fusion method or BP reaction to construct them. With the In-Fusion method, we used a promoter sequence to substitute for promoter fragment in an entry clone A used in LR reactions in Three-Fragment Multisite gateway system. In brief, a linearized vector without promoter sequence was prepared by reverse PCR amplification, using an entry clone A as a template. The promoter was PCR amplified from wild-type N2 genomic DNA, using primers containing about 20 bp sequences of the promoter and homologous sequences (15–20 bp) of the linearized vector. The linearized vector and amplified promoter fragment were recombined to generate entry clone A, using ClonExpress®II One Step Cloning Kit (Vazyme Biotech Co.).

Construction of an entry clone B containing a gene of interest. A large part of entry clone B containing a gene of interest (or gene vector) was constructed by BP recombination reactions. A small part of the gene vectors was constructed by the In-Fusion method used for constructing the expression plasmids containing receptors used in this study.

An entry clone B contributing sequences of slot 2 in the expression plasmid contains sequence of a target gene. The most “B” entry clones were constructed using pDONR™ 221 donor vector (Invitrogen) through BP reactions. All the DNA sequences used to construct entry clones B involved in this project were amplified by PCR with the primers containing attB1 and attB2 recombination sites.

Construction of an entry clone C. An entry clone C contributing sequences of slot 3 in the expression plasmids contains a sequence of unc-54 3’ UTR, sl2d::GFP, or sl2e::TagRFP-t. The “C” entry clones were constructed by use of pDONR-P2R-P3 donor vector through standard BP reactions. The corresponding PCR products with attB2R and attB3 sites were amplified by PCR. The “C” entry clone containing unc-54 3’ UTR, sl2d::GFP, or sl2e::TagRFP, were used to construct expression plasmids. All primers and related information for cloning these promoters and genes are listed in supplementary table S3

Behavioral assays

Mock and salt conditioning. All behavioral experiments were conducted using the synchronized day-1 adult worms. The chemotaxis to NaCl and salt chemotaxis learning of mock- and NaCl/food-deprivation association-conditioned (salt-conditioned, hereafter) animals were performed on test agar plates. For NaCl/food-deprivation association conditioning or salt conditioning, worms on culture plates were washed out with 100 mM NaCl-CTX solution (100 mM NaCl, 5 mM KH₂PO₄/K₂HPO₄ with pH 6.6, 1 mM CaCl₂, 1 mM MgSO₄) and transferred into a 1.5 mL centrifugal tube. The worms were washed three times with about 1 mL 100 mM NaCl-CTX solution for about 5 min to remove bacteria from worm’s body. Then, the worms were soaked in 1.2 mL 100 mM NaCl-CTX solution for about 14 min. Lastly, the worms were washed with CTX buffer without NaCl one time to remove NaCl. Total time for these manipulations was about 20 min. For mock conditioning, the worms were treated with CTX buffer (5 mM KH₂PO₄/K₂HPO₄ with pH 6.6, 1 mM CaCl₂, 1 mM MgSO₄) using the same procedures and time as used for the salt conditioning.

Preparing test agar plates. A test of chemotaxis to NaCl or salt chemotaxis learning (SCL) was performed on a test NGM plate in diameter of 9 cm in a Petri dish at room temperature of 20°C. The test plates contain two chemotaxis test regions without (region A) or with (region B) 20 mM NaCl in opposite direction. They were prepared with the following protocol. 15mL hot agar solution (5 mM KH₂PO₄/K₂HPO₄ with pH 6.6, 1 mM CaCl₂, 1 mM MgSO₄, 2% agar) at about 80°C was poured into a 9 cm diameter Petri dish. After naturally cooled and solidified, two circular test regions in 3.5 cm diameter close to the edge of dish in the opposite direction were made. The agar in the test regions was cut with a 3.5 cm dish and then removed. The test regions were refilled with a hot agar solution with 20 mM NaCl (20 mM NaCl, 5 mM KH₂PO₄/K₂HPO₄ with pH 6.6, 1 mM CaCl₂, 1 mM MgSO₄, 2% agar) or without NaCl (5 mM KH₂PO₄/K₂HPO₄ with pH 6.6, 1 mM CaCl₂, 1 mM MgSO₄, 2% agar), respectively. The volume of refilling agar solution was carefully controlled to make sure that whole plate was even. Lastly, the test plates were incubated at 37°C for 3 hours and then left at room temperature for 10 hours.

Chemotaxis and salt chemotaxis learning assay. The mock- or salt-conditioned animals of number about 200 in about 20 µL solution were transferred onto the center of test plate. The residual solution was absorbed with tissue. Let worms freely move on the test plate for 10 minutes. Then, a circle of 0.5 mM sodium azide was made around the edge of two test regions to paralyze worms. The number of worms in the test regions were counted under a stereomicroscope. The index of chemotaxis to NaCl in mock-conditioned animals was calculated as (Number_B - Number_A) / (Number_B + Number_A). Here, Number_A, the number of worms within the test region without NaCl; Number_B, the number of animals within the test region with 20 mM NaCl. The learning index was calculated as averaged chemotaxis index of mock-conditioned animals minus that of salt-conditioned worms in a test. All conditioning and chemotaxis tests were performed at 20°C.

Genetic and chemogenetic manipulations of tested neurons

Genetic manipulations of tested neurons. We employed neuron-specific extrachromosomal expression of TeTx to genetically block vesicle fusion with the plasma membrane and thus neurotransmitter release in the tested neurons. TeTx is a specific protease of synaptobrevin⁷⁰. It is used successfully to inhibit the chemical neurotransmission in tested neurons in C. elegans^{72, 73, 77, 105}.

Chemogenetic manipulations of tested neurons. To tempo-spatially silence the tested neurons, we used chemogenetic neuronal manipulation. In short, we used neuron type-specific promoters to drive expression of the Drosophila histamine-gated chloride channel subunit HisCl1 gene¹⁰⁵ in the tested neurons and employed 10 mM histamine to activate the chloride ion channel^{72–74, 77, 105}. To drive TeTx or HisCl1 expression in ASER, RIC, and RIS, the promoters gcy-5p (3.2 kb, in ASER), tbh-1p (4.5 kb, in RIC), and frpr-10p (2 kb, in RIS) were used, respectively. HisCl1 channel may have a low probability of opening without ligand histamine, causing a leaking current in some cases. To reduce the leaky effect on tests, we injected the amount of the HisCl1 plasmids by reducing its concentrations to 15 ng/uL.

For chemogenetic inhibition of the tested neurons in learning stage, we conditioned animals with 10 mM histamine in solutions used in mock- and salt-conditioning. Then, we tested the treated animals’ chemotaxis on normal test plates.

For chemogenetic inhibition of the tested neurons in memory recalling stage, we used a test agar plate containing10 mM histamine to assay the normally conditioned animals’ chemotaxis behavior. These test plates were prepared with the same procedures for preparation of the normal test plates, except all agar solutions contained10 mM histamine. To prepare the agar solutions containing 10 mM histamine, a stock solution of histamine (1 M in CTX buffer) was diluted with ultrapure water into a 100 mM working solution, and 10% (v/v) working solution was added into agar solutions at approximately 60°C before preparing the plates.

Calcium imaging

The calcium (Ca²⁺) responses in neuronal soma were measured by detecting changes in the fluorescence intensity of the genetically encoded Ca²⁺ indicator R-GECO1 and G-CaMP3.0. A home-made microfluidic device was used for calcium imaging, as previously described^{72–75, 77}. Briefly, a mock- or salt-conditioned worm was loaded into the worm channel of the microfluidic chip with its nose exposed to solutions under laminar flow. For each Ca²⁺ assay with duration of 80 s in ASER, RIS, RIC, and AIY, NaCl stimulation was performed by a switch between CTX buffer and 50 mM NaCl-CTX solution: CTX for 20 s-NaCl-CTX for 30 s-CTX for 30 s. The osmolarities of all solutions used for Ca²⁺ imaging were adjusted to 350 mOsmM with sucrose. The solutions were delivered via programmable automatic drug-feeding equipment (MPS-2, InBio Life Science Instrument Co. Ltd, Wuhan, Hubei Province, China). R-GECO1.0 was excited by 525–530 nm light emitted by an Osram Diamond Dragon LTW5AP light-emitting diode (LED) model (Osram, Marcel-Breuer-Strasse 6, Munich, Germany) mounted to a multi-LED light source (MLS102, InBio Life Science Instrument Co. Ltd) and filtered by a Semrock FF01-593/40 − 25 emission filter (IDEX Health & Science, LLC, Oak Harbor, WA, USA) under an Olympus IX-70 inverted microscope (Olympus) equipped with a 40× objective lens (numerical aperture (NA) = 1.3, Carl Zeiss MicroImaging GmhH, Göttingen, Germany). Fluorescence images were captured by an Andor iXonEM + DU885K EMCCD camera with 256 × 256 pixels at 10 frames per second. Each animal was exposed under fluorescent excitation light for 30 s before recording to reduce the light-evoked calcium transients and was recorded once only. The averaged fluorescence intensity in each frame of the region of interest (ROI) of the soma was captured and analyzed by using Image-Pro Plus 6 (Media Cybernetics, Inc., Rockville, MD, USA). The average signal of an adjacent ROI in all frames was used to subtract the background. The average fluorescence intensity within the initial 20 s before stimulation was taken as basal signal F₀. The changes in fluorescence intensity relative to the initial intensity F₀, ΔF = (F - F₀) / F₀, were plotted as a function of time for all curves. The Ca²⁺ signals, ΔF, were displayed as the mean values in various colors and SEM in light grey using IGOR Pro 6.10 (WaveMetrics, Inc., Lake Oswego, OR, USA), as heatmap using Matlab-R2019a (MathWorks), or boxplot using GraphPad Prism 8 (GraphPad Software).

Confocal fluorescence imaging

All confocal fluorescence imaging was performed using an Andor Revolution XD laser confocal microscope system (Andor Technology plc., Springvale Business Park, Belfast, UK) based on a spinning-disk confocal scanning head CSU-X1 (Yokogawa Electric Corporation, Musashino-shi, Tokyo, Japan). The confocal system was mounted on an Olympus IX-71 inverted microscope (Olympus, Tokyo, Japan) and controlled by Andor IQ 1.91 software. Live young adult worms were mounted on a 2% (w/v) agarose pad with 10 mM levamisole (Sigma-Aldrich). All fluorescent images were imaged by ×60 objective lens (numerical aperture = 1.45, Olympus) and captured by an Andor iXonEM + DU-897D EMCCD camera. The images were displayed using Image J 1.52o software (Wayne Rasband, National Institutes of Health, USA).

Locomotion analysis

Day-1 young adult worms cultured by standard methods were transferred onto an area in NGM plates (diameter 6 cm) with a thin layer of OP50 and allowed them to move freely for 3 min. Worm locomotion was recorded under a Zeiss Discovery V8 stereomicroscope (Carl Zeiss MicroImaging GmbH, Göttingen, Germany). The image sequences were captured with an Andor iXonEM + DV885K EMCCD camera (Andor Technology plc., Springvale Business Park, Belfast, UK) at 10 frame per second. The worm locomotion speed was analyzed by use of Worm Lab software.

Neuropeptide-receptor interaction prediction with Alphafold2

The receptor structure is first modeled independently using AlphaFold2 and then input into the protein template library. To model the interaction between the neuropeptide and the receptor, link the two sequences with 30 glycines, and use AlphaFold2 to predict the structure of the neuropeptide-receptor complex. Parameter settings: iteration numbers (3, 6, 9, 12, 15, 18, 20) and randomly set 10 seed numbers. Among the five models, as long as one generates a distance of less than 8 Å between any Cβ atoms of the neuropeptide and the receptor, and the binding site of the neuropeptide is located at the N-terminus of the receptor, it is considered valid. The hardware used includes 4 NVIDIA A40 GPUs, and the MSA search employs default parameters⁶⁶.

FRPR-10 expression in X. laevis oocytes and whole-cell current recording

FRPR-10 expression in X. laevis oocytes. frpr-10 cDNAs were flanked with BamHI and HindIII sites by PCR and cloned into a pGH19 vector with by T4 ligase. The cRNAs of frpr-10, GIRK1, and GIRK4, were prepared using an mMachine™ T7 kit (Thermo Fisher scientific, Waltham, MA, USA). The sense of frpr-10 (50 ng), GIRK1 (0.5 ng), and GIRK4 (0.5 ng), were injected into a X. laevis oocyte. The cRNA-injected oocytes were incubated at 18°C in ND96 medium (96 mM NaCl, 2.5 mM KCl, 1 mM MgCl₂ and 5 mM HEPES, pH 7.6) for 2–5 d before recording.

Whole-cell current recording. Whole-cell current was made using the two-electrode voltage-clamp technique at a holding potential of − 80 mV as described^{68, 69}. The recording chamber was continuously perfused with ND96 medium. To assay activation of GIRK channels, we equilibrated oocytes in high-K⁺ medium (96 mM KCl and 2.5 mM NaCl instead of 2.5 mM KCl and 96 mM NaCl) to reverse the K⁺ gradient and measured inward currents before, during and after addition of test peptide. Data were acquired with Clampex 8.0 software (Molecular Devices) and analyzed offline with Clampfit (Molecular Devices). Peptides were synthesized by the Guoping Pharmaceutical Company (Hefei, Anhui Province, China).

Data analyses

Data of chemotaxis and learning index are displayed as box plots, with each dot representing the data from each independent test. The Ca²⁺ signal data are expressed as heatmaps, box plots, or as the means ± SEM indicated by solid traces ± gray shading. Data were statistically analyzed using software packages in GraphPad Prism 8 (GraphPad Software, Inc., San Diego, CA, USA). When the comparison was limited to 2 groups, a two-tailed t test with unpaired samples was used to analyze differences and calculate p-values. When more than two groups of data were compared, data were analyzed by ordinary one-way or two-way analysis of variance (ANOVA), with recommended post hoc tests in the GraphPad Prism 8 software package. Dunnett’s multiple comparison correction was applied when multiple samples were compared to a single sample, i.e., wild-type N2 or other controls. Tukey’s multiple comparison correction was used when multiple samples were compared. The p-value is indicated as follows: ns, not significant, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, and in different colors for varied comparisons.

Declaration of interests

The authors declare no competing interests.

Data availability

The authors declare that data supporting the findings of this study are available within the paper and its Supplementary Information files. All additional information and the C. elegans mutants generated in this study will be made available upon reasonable request to the corresponding author.

Kandel ER. The molecular biology of memory: cAMP, PKA, CRE, CREB-1, CREB-2, and CPEB. Mol Brain 5, 14 (2012).
Kandel ER, LeDoux J. Cellular mechanisms of implicit memory storage and the biological basis of individuality. In: Principles of Neuronal Science (eds Kandel ER, Koester JD, Mack SH, Siegelbaum SA). McGraw Hill (2021).
Owen GR, Brenner EA. Mapping molecular memory: Navigating the cellular pathways of learning. Cell Mol Neurobiol 32, 919-941 (2012).
Gluck MA, Mercado E, Myers CE. Learning and memory: From brain to behavior, 4th edn. Worth Publishers (2020).
Clark RE, Martin SJ. Behavioral neuroscience of learning and memory. Springer (2018).
Ardiel EL, Rankin CH. Behavioral plasticity in the C. elegans mechanosensory circuit. J Neurogenet 22, 239-255 (2008).
Sakata JT, Catalano I, Woolley SC. Mechanisms, development, and comparative perspectives on experience-dependent plasticity in social behavior. J Exp Zool A Ecol Integr Physiol 337, 35-49 (2022).
Ishihara T, et al. HEN-1, a secretory protein with an LDL receptor motif, regulates sensory integration and learning in. Cell 109, 639-649 (2002).
Sanchez-Alcaniz JA, Benton R. Multisensory neural integration of chemical and mechanical signals. Bioessays 39, (2017).
Bailey CH, Chen M. Long-term sensitization in Aplysia increases the number of presynaptic contacts onto the identified gill motor neuron L7. Proc Natl Acad Sci U S A 85, 9356-9359 (1988).
Bailey CH, Kandel ER. Synaptic remodeling, synaptic growth and the storage of long-term memory in Aplysia. Prog Brain Res 169, 179-198 (2008).
Menzel R, Benjamin PR. Invertebrate Learning and Memory. In: Handbook of Behavioral Neuroscience (eds Menzel R, Benjamin PR). Elsevier (2013).
McMillen A, Chew YL. Neural mechanisms of dopamine function in learning and memory in Caenorhabditis elegans. Neuronal Signal 8, NS20230057 (2024).
Rahmani A, Chew YL. Investigating the molecular mechanisms of learning and memory using Caenorhabditis elegans. J Neurochem 159, 417-451 (2021).
Sasakura H, Mori I. Behavioral plasticity, learning, and memory in C. elegans. Curr Opin Neurobiol 23, 92-99 (2013).
Bargmann CI, Hartwieg E, Horvitz HR. Odorant-selective genes and neurons mediate olfaction in C. elegans. Cell 74, 515-527 (1993).
Ward S. Chemotaxis by the nematode Caenorhabditis elegans: identification of attractants and analysis of the response by use of mutants. Proc Natl Acad Sci U S A 70, 817-821 (1973).
Hukema RK, Rademakers S, Dekkers MP, Burghoorn J, Jansen G. Antagonistic sensory cues generate gustatory plasticity in Caenorhabditis elegans. EMBO J 25, 312-322 (2006).
Saeki S, Yamamoto M, Iino Y. Plasticity of chemotaxis revealed by paired presentation of a chemoattractant and starvation in the nematode Caenorhabditis elegans. J Exp Biol 204, 1757-1764 (2001).
Tomioka M, Adachi T, Suzuki H, Kunitomo H, Schafer WR, Iino Y. The insulin/PI₃-kinase pathway regulates salt chemotaxis learning in Caenorhabditis elegans. Neuron 51, 613-625 (2006).
Luo L, et al. Dynamic encoding of perception, memory, and movement in a C. elegans chemotaxis circuit. Neuron 82, 1115-1128 (2014).
Niu S, Zhou W, Li Y, Huang X. The signaling pathway of levamisole-sensitive-acetylcholine receptors involved in short-term forgetting of Caenorhabditis elegans. Mol Genet Genomics 297, 1027-1038 (2022).
Moro CA, et al. Adenylosuccinate lyase deficiency affects neurobehavior via perturbations to tyramine signaling in Caenorhabditis elegans. PLoS Genet 19, e1010974 (2023).
Rosikon KD, Bone MC, Lawal HO. Regulation and modulation of biogenic amine neurotransmission in Drosophila and Caenorhabditis elegans. Front Physiol 14, 970405 (2023).
Beck B, Pourie G. Ghrelin, neuropeptide Y, and other feeding-regulatory peptides active in the hippocampus: Role in learning and memory. Nutr Rev 71, 541-561 (2013).
Bowers ME, Choi DC, Ressler KJ. Neuropeptide regulation of fear and anxiety: Implications of cholecystokinin, endogenous opioids, and neuropeptide Y. Physiol Behav 107, 699-710 (2012).
Dyzma M, Boudjeltia KZ, Faraut B, Kerkhofs M. Neuropeptide Y and sleep. Sleep Med Rev 14, 161-165 (2010).
Feany MB. Neuropeptide modulation of learning and memory processes. Rev Neurosci 7, 151-164 (1996).
Gilpin NW, Roberto M. Neuropeptide modulation of central amygdala neuroplasticity is a key mediator of alcohol dependence. Neurosci Biobehav Rev 36, 873-888 (2012).
Morley JE. Neuropeptide regulation of appetite and weight. Endocr Rev 8, 256-287 (1987).
Zager EL, Black PM. Neuropeptides in human memory and learning processes. Neurosurgery 17, 355-369 (1985).
Peymen K, Watteyne J, Borghgraef C, Van Sinay E, Beets I, Schoofs L. Myoinhibitory peptide signaling modulates aversive gustatory learning in Caenorhabditis elegans. PLoS Genet 15, e1007945 (2019).
Van Bael S, et al. Mass spectrometric evidence for neuropeptide-amidating enzymes in Caenorhabditis elegans. J Biol Chem 293, 6052-6063 (2018).
Van Bael S, Zels S, Boonen K, Beets I, Schoofs L, Temmerman L. A Caenorhabditis elegans mass spectrometric resource for neuropeptidomics. J Am Soc Mass Spectrom 29, 879-889 (2018).
Li WQ, Kennedy SG, Ruvkun G. daf-28 encodes a C. elegans insulin superfamily member that is regulated by environmental cues and acts in the DAF-2 signaling pathway. Gene Dev 17, 844-858 (2003).
Pierce SB, et al. Regulation of DAF-2 receptor signaling by human insulin and ins-1, a member of the unusually large and diverse C. elegans insulin gene family. Genes Dev 15, 672-686 (2001).
Li C, Kim K, Nelson LS. FMRFamide-related neuropeptide gene family in Caenorhabditis elegans. Brain Res 848, 26-34 (1999).
Frooninckx L, et al. Neuropeptide GPCRs in C. elegans. Front Endocrinol (Lausanne) 3, 167 (2012).
Nathoo AN, Moeller RA, Westlund BA, Hart AC. Identification of neuropeptide-like protein gene families in Caenorhabditis elegans and other species. Proc Natl Acad Sci U S A 98, 14000-14005 (2001).
Beets I, et al. Vasopressin/oxytocin-related signaling regulates gustatory associative learning in C. elegans. Science 338, 543-545 (2012).
Chen Z, Hendricks M, Cornils A, Maier W, Alcedo J, Zhang Y. Two insulin-like peptides antagonistically regulate aversive olfactory learning in C. elegans. Neuron 77, 572-585 (2013).
Kauffman AL, Ashraf JM, Corces-Zimmerman MR, Landis JN, Murphy CT. Insulin signaling and dietary restriction differentially influence the decline of learning and memory with age. PLoS Biol 8, e1000372 (2010).
Lee K, Mylonakis E. An Intestine-Derived Neuropeptide Controls Avoidance Behavior in Caenorhabditis elegans. Cell Rep 20, 2501-2512 (2017).
Lin CH, Tomioka M, Pereira S, Sellings L, Iino Y, van der Kooy D. Insulin signaling plays a dual role in Caenorhabditis elegans memory acquisition and memory retrieval. J Neurosci 30, 8001-8011 (2010).
Watteyne J, et al. Neuromedin U signaling regulates retrieval of learned salt avoidance in a C. elegans gustatory circuit. Nat Commun 11, 2076 (2020).
Wu T, et al. Pheromones modulate learning by regulating the balanced signals of two insulin-like peptides. Neuron 104, 1095-1109 e1095 (2019).
Yamada K, et al. Olfactory plasticity is regulated by pheromonal signaling in Caenorhabditis elegans. Science 329, 1647-1650 (2010).
Koelle MR. Neurotransmitter signaling through heterotrimeric G proteins: insights from studies in C. elegans. WormBook 2018, 1-52 (2018).
Jekely G. Global view of the evolution and diversity of metazoan neuropeptide signaling. Proc Natl Acad Sci U S A 110, 8702-8707 (2013).
Mirabeau O, Joly JS. Molecular evolution of peptidergic signaling systems in bilaterians. Proc Natl Acad Sci U S A 110, E2028-2037 (2013).
Van Sinay E, et al. Evolutionarily conserved TRH neuropeptide pathway regulates growth in Caenorhabditis elegans. Proc Natl Acad Sci U S A 114, E4065-E4074 (2017).
Chang LW, Perez A. Ranking Peptide Binders by Affinity with AlphaFold. Angew Chem Int Edit 62, (2023).
Humphreys IR, et al. Computed structures of core eukaryotic protein complexes. Science 374, eabm4805 (2021).
Lin Z, et al. Evolutionary-scale prediction of atomic-level protein structure with a language model. Science 379, 1123-1130 (2023).
Tsuchiya Y, Yamamori Y, Tomii K. Protein-protein interaction prediction methods: from docking-based to AI-based approaches. Biophys Rev 14, 1341-1348 (2022).
Wu R, et al. High-resolution de novo structure prediction from primary sequence. BioRxiv, 2022.2007. 2021.500999 (2022).
Jumper J, et al. Highly accurate protein structure prediction with AlphaFold. Nature 596, 583-589 (2021).
Nussinov R, Zhang M, Liu Y, Jang H. AlphaFold, Artificial Intelligence (AI), and Allostery. J Phys Chem B 126, 6372-6383 (2022).
Wicks SR, de Vries CJ, van Luenen HGAM, Plasterk RHA. CHE-3, a cytosolic dynein heavy chain, is required for sensory cilia structure and function in Caenorhabditis elegans. Developmental Biology 221, 295-307 (2000).
Hukema RK, Rademakers S, Jansen G. Gustatory plasticity in C. elegans involves integration of negative cues and NaCl taste mediated by serotonin, dopamine, and glutamate. Learn Mem 15, 829-836 (2008).
Kass J, Jacob TC, Kim P, Kaplan JM. The EGL-3 proprotein convertase regulates mechanosensory responses of Caenorhabditis elegans. J Neurosci 21, 9265-9272 (2001).
Gershkovich MM, Gross VE, Kaiser A, Promel S. Pharmacological and functional similarities of the human neuropeptide Y system in C. elegans challenges phylogenetic views on the FLP/NPR system. Cell Commun Signal 17, 123 (2019).
Marques F, Falquet L, Vandewyer E, Beets I, Glauser DA. Signaling via the FLP-14/FRPR-19 neuropeptide pathway sustains nociceptive response to repeated noxious stimuli in C. elegans. PLoS Genet 17, e1009880 (2021).
Bryant P, Pozzati G, Elofsson A. Improved prediction of protein-protein interactions using AlphaFold2. Nat Commun 13, 1265 (2022).
Danneskiold-Samsøe NB, et al. Rapid and accurate deorphanization of ligand–receptor pairs using AlphaFold. bioRxiv, (2023).
Tsaban T, Varga JK, Avraham O, Ben-Aharon Z, Khramushin A, Schueler-Furman O. Harnessing protein folding neural networks for peptide–protein docking. Nat Commun 13, 176 (2022).
Kishore R, et al. Automated generation of gene summaries at the Alliance of Genome Resources. Database-Oxford, (2020).
Ringstad N, Horvitz HR. FMRFamide neuropeptides and acetylcholine synergistically inhibit egg-laying by C. elegans. Nat Neurosci 11, 1168-1176 (2008).
Rogers C, et al. Inhibition of Caenorhabditis elegans social feeding by FMRFamide-related peptide activation of NPR-1. Nat Neurosci 6, 1178-1185 (2003).
Schiavo G, et al. Tetanus and botulinum-B neurotoxins block neurotransmitter release by proteolytic cleavage of synaptobrevin. Nature 359, 832-835 (1992).
Guo M, et al. Dissecting molecular and circuit mechanisms for inhibition and delayed response of ASI neurons during nociceptive stimulus. Cell Rep 25, 1885-1897 e1889 (2018).
Guo M, et al. Reciprocal inhibition between sensory ASH and ASI neurons modulates nociception and avoidance in Caenorhabditis elegans. Nat Commun 6, 5655 (2015).
Li R, et al. GCY-20 signaling controls suppression of Caenorhabditis elegans egg laying by moderate cold. Cell Rep 43, 113708 (2024).
Liu H, Qin LW, Li R, Zhang C, Al-Sheikh U, Wu ZX. Reciprocal modulation of 5-HT and octopamine regulates pumping via feedforward and feedback circuits in C. elegans. Proc Natl Acad Sci U S A 116, 7107-7112 (2019).
Liu H, et al. Disexcitation in the ASH/RIM/ADL negative feedback circuit fine-tunes hyperosmotic sensation and avoidance in Caenorhabditis elegans. Front Mol Neurosci 16, 1101628 (2023).
Macosko EZ, et al. A hub-and-spoke circuit drives pheromone attraction and social behaviour in C. elegans. Nature 458, 1171-1175 (2009).
Wen X, et al. Signal decoding for glutamate modulating egg laying oppositely in Caenorhabditis elegans under varied environmental conditions. Iscience 23, 101588 (2020).
Bastiani C, Mendel J. Heterotrimeric G proteins in C. elegans. WormBook, 1-25 (2006).
Gereau GB, Garrison SD, McElligott ZA. Neurotensin and energy balance. J Neurochem 166, 189-200 (2023).
Erwin VG, Jones BC. Genetic correlations among ethanol-related behaviors and neurotensin receptors in long sleep (LS) x short sleep (SS) recombinant inbred strains of mice. Behav Genet 23, 191-196 (1993).
Lénárd L, et al. Substance P and neurotensin in the limbic system: Their roles in reinforcement and memory consolidation. Neurosci Biobehav Rev 85, 1-20 (2018).
Li J, et al. Neurotensin receptor 1 gene (NTSR1) polymorphism is associated with working memory. Plos One 6, e17365 (2011).
Osadchii OE. Emerging role of neurotensin in regulation of the cardiovascular system. Eur J Pharmacol 762, 184-192 (2015).
Cook SJ, et al. Whole-animal connectomes of both Caenorhabditis elegans sexes. Nature 571, 63-71 (2019).
White JG, Southgate E, Thomson JN, Brenner S. The structure of the nervous system of the nematode Caenorhabditis elegans. Philos Trans R Soc Lond B Biol Sci 314, 1-340 (1986).
Altun ZF, Chen B, Wang ZW, Hall DH. High resolution map of Caenorhabditis elegans gap junction proteins. Dev Dyn 238, 1936-1950 (2009).
Durham TJ, et al. Comprehensive characterization of tissue-specific chromatin accessibility in L2 Caenorhabditis elegans nematodes. Genome Res 31, 1952-1969 (2021).
Jin Y JE, Hartwieg E, Horvitz HR. The Caenorhabditis elegans gene unc-25 encodes glutamic acid decarboxylase and is required for synaptic transmission but not synaptic development. J Neurol Sci, (1999).
Janssen T, et al. Discovery and characterization of a conserved pigment dispersing factor-like neuropeptide pathway in Caenorhabditis elegans. J Neurochem 111, 228-241 (2009).
Lockhead D, et al. The tubulin repertoire of Caenorhabditis elegans sensory neurons and its context‑dependent role in process outgrowth. Molecular biology of the cell 27, 3717-3728 (2016).
Meelkop E, et al. PDF receptor signaling in Caenorhabditis elegans modulates locomotion and egg-laying. Mol Cell Endocrinol 361, 232-240 (2012).
Greer ER, Perez CL, Van Gilst MR, Lee BH, Ashrafi K. Neural and molecular dissection of a C. elegans sensory circuit that regulates fat and feeding. Cell Metab 8, 118-131 (2008).
Lemieux GA, Cunningham KA, Lin L, Mayer F, Werb Z, Ashrafi K. Kynurenic acid is a nutritional cue that enables behavioral plasticity. Cell 160, 119-131 (2015).
Li Z, et al. Dissecting a central flip-flop circuit that integrates contradictory sensory cues in C. elegans feeding regulation. Nat Commun 3, 776 (2012).
Liao CP, Chiang YC, Tam WH, Chen YJ, Chou SH, Pan CL. Neurophysiological basis of stress-induced aversive memory in the nematode Caenorhabditis elegans. Curr Biol 32, 5309-5322 e5306 (2022).
Liu M, et al. Eleutheroside E enhances the long-term memory of radiation-damaged C. elegans through G-protein-coupled receptor and neuropeptide signaling pathways. J Nat Prod 83, 3315-3323 (2020).
Noble T, Stieglitz J, Srinivasan S. An integrated serotonin and octopamine neuronal circuit directs the release of an endocrine signal to control C. elegans body fat. Cell Metab 18, 672-684 (2013).
Tao J, Ma YC, Yang ZS, Zou CG, Zhang KQ. Octopamine connects nutrient cues to lipid metabolism upon nutrient deprivation. Sci Adv 2, (2016).
Suo S, Culotti JG, Van Tol HH. Dopamine counteracts octopamine signalling in a neural circuit mediating food response in C. elegans. EMBO J 28, 2437-2448 (2009).
Suo S, Kimura Y, Van Tol HH. Starvation induces cAMP response element-binding protein-dependent gene expression through octopamine-Gq signaling in Caenorhabditis elegans. J Neurosci 26, 10082-10090 (2006).
Roeder T. Octopamine in invertebrates. Prog Neurobiol 59, 533-561 (1999).
Roeder T. Tyramine and octopamine: Ruling behavior and metabolism. Annu Rev Entomol 50, 447-477 (2005).
Tully K, Li Y, Tsvetkov E, Bolshakov VY. Norepinephrine enables the induction of associative long-term potentiation at thalamo-amygdala synapses. Proc Natl Acad Sci U S A 104, 14146-14150 (2007).
Alkema MJ, Hunter-Ensor M, Ringstad N, Horvitz HR. Tyramine functions independently of octopamine in the Caenorhabditis elegans nervous system. Neuron 46, 247-260 (2005).
Pokala N, Liu Q, Gordus A, Bargmann CI. Inducible and titratable silencing of Caenorhabditis elegans neurons in vivo with histamine-gated chloride channels. Proc Natl Acad Sci U S A 111, 2770-2775 (2014).
Wu JJ, et al. Positive interaction between ASH and ASK sensory neurons accelerates nociception and inhibits behavioral adaptation. Iscience 25, 105287 (2022).
Bargmann CI. Chemosensation in C. elegans. WormBook, 1-29 (2006).
Pierce-Shimomura JT, Faumont S, Gaston MR, Pearson BJ, Lockery SR. The homeobox gene lim-6 is required for distinct chemosensory representations in C. elegans. Nature 410, 694-698 (2001).
Suzuki H, Thiele TR, Faumont S, Ezcurra M, Lockery SR, Schafer WR. Functional asymmetry in Caenorhabditis elegans taste neurons and its computational role in chemotaxis. Nature 454, 114-117 (2008).
Adachi T, et al. Reversal of salt preference is directed by the insulin/PI3K and G_q/PKC signaling in Caenorhabditis elegans. Genetics 186, 1309-1319 (2010).
Oda S, Tomioka M, Iino Y. Neuronal plasticity regulated by the insulin-like signaling pathway underlies salt chemotaxis learning in Caenorhabditis elegans. J Neurophysiol 106, 301-308 (2011).
Ohno H, Kato S, Naito Y, Kunitomo H, Tomioka M, Iino Y. Role of synaptic phosphatidylinositol 3-kinase in a behavioral learning response in C. elegans. Science 345, 313-317 (2014).
Tomioka M, Naito Y, Kuroyanagi H, Iino Y. Splicing factors control C. elegans behavioural learning in a single neuron by producing DAF-2c receptor. Nat Commun 7, 11645 (2016).
Lim JP, et al. Loss of CaMKI function disrupts salt aversive learning in C. elegans. J Neurosci 38, 6114-6129 (2018).
Iino Y, Yoshida K. Parallel use of two behavioral mechanisms for chemotaxis in Caenorhabditis elegans. J Neurosci 29, 5370-5380 (2009).
Pierce-Shimomura JT, Morse TM, Lockery SR. The fundamental role of pirouettes in Caenorhabditis elegans chemotaxis. J Neurosci 19, 9557-9569 (1999).
Kimura KD, Fujita K, Katsura I. Enhancement of odor avoidance regulated by dopamine signaling in Caenorhabditis elegans. J Neurosci 30, 16365-16375 (2010).
Li Z, Liu J, Zheng M, Xu XZ. Encoding of both analog- and digital-like behavioral outputs by one C. elegans interneuron. Cell 159, 751-765 (2014).
Debnath A, Williams PDE, Bamber BA. Reduced Ca²⁺ transient amplitudes may signify increased or decreased depolarization depending on the neuromodulatory signaling pathway. Front Neurosci 16, 931328 (2022).
O'Donnell MP, Fox BW, Chao PH, Schroeder FC, Sengupta P. A neurotransmitter produced by gut bacteria modulates host sensory behaviour. Nature 583, 415-420 (2020).
Rex E, Hapiak V, Hobson R, Smith K, Xiao H, Komuniecki R. TYRA-2 (F01E11.5): A Caenorhabditis elegans tyramine receptor expressed in the MC and NSM pharyngeal neurons. J Neurochem 94, 181-191 (2005).
Ghosh DD, et al. Neural architecture of hunger-dependent multisensory decision making in C. elegans. Neuron 92, 1049-1062 (2016).
Fu J, et al. AIM interneurons mediate feeding suppression through the TYRA-2 receptor in C. elegans. Biophys Rep 4, 17-24 (2018).
Chute CD, et al. Co-option of neurotransmitter signaling for inter-organismal communication in C. elegans. Nat Commun 10, 3186 (2019).
Laszlo BR, et al. The role of intra-amygdaloid neurotensin and dopamine interaction in spatial learning and memory. Biomedicines 10, (2022).
Laszlo K, Toth K, Kertes E, Peczely L, Ollmann T, Lenard L. Effects of neurotensin in amygdaloid spatial learning mechanisms. Behav Brain Res 210, 280-283 (2010).
Li H, et al. Neurotensin orchestrates valence assignment in the amygdala. Nature 608, 586-592 (2022).
Xiao Z, et al. Activation of neurotensin receptor 1 facilitates neuronal excitability and spatial learning and memory in the entorhinal cortex: beneficial actions in an Alzheimer's disease model. J Neurosci 34, 7027-7042 (2014).
Borbely E, Scheich B, Helyes Z. Neuropeptides in learning and memory. Neuropeptides 47, 439-450 (2013).
Del Pino I, et al. Abnormal wiring of CCK(+) basket cells disrupts spatial information coding. Nat Neurosci 20, 784-792 (2017).
Krashes MJ, DasGupta S, Vreede A, White B, Armstrong JD, Waddell S. A neural circuit mechanism integrating motivational state with memory expression in Drosophila. Cell 139, 416-427 (2009).
Ma S, Hangya B, Leonard CS, Wisden W, Gundlach AL. Dual-transmitter systems regulating arousal, attention, learning and memory. Neurosci Biobehav Rev 85, 21-33 (2018).
Marquina-Solis J, et al. Antagonism between neuropeptides and monoamines in a distributed circuit for pathogen avoidance. Cell Rep 43, (2024).
Rangel Guerrero DK, et al. Hippocampal cholecystokinin-expressing interneurons regulate temporal coding and contextual learning. Neuron 112, 2045-2061 e2010 (2024).
Varas MM, Perez MF, Ramirez OA, de Barioglio SR. Increased susceptibility to LTP generation and changes in NMDA-NR1 and -NR2B subunits mRNA expression in rat hippocampus after MCH administration. Peptides 24, 1403-1411 (2003).
Wang C, et al. Neuropeptide S displays as a key neuromodulator in olfactory spatial memory. Chem Senses 45, 195-202 (2020).
Burke DF, et al. Towards a structurally resolved human protein interaction network. Nat Struct Mol Biol 30, 216-225 (2023).
Turek M, Besseling J, Spies JP, Konig S, Bringmann H. Sleep-active neuron specification and sleep induction require FLP-11 neuropeptides to systemically induce sleep. Elife 5, (2016).
Turek M, Lewandrowski I, Bringmann H. An AP2 transcription factor is required for a sleep-active neuron to induce sleep-like quiescence in C. elegans. Curr Biol 23, 2215-2223 (2013).
Busack I, Bringmann H. A sleep-active neuron can promote survival while sleep behavior is disturbed. PLoS Genet 19, e1010665 (2023).
Steuer Costa W, et al. A GABAergic and peptidergic sleep neuron as a locomotion stop neuron with compartmentalized Ca²⁺ dynamics. Nat Commun 10, 4095 (2019).
Rasch B, Born J. About sleep's role in memory. Physiol Rev 93, 681-766 (2013).
Diekelmann S, Born J. The memory function of sleep. Nat Rev Neurosci 11, 114-126 (2010).
Dissel S, Melnattur K, Shaw PJ. Sleep, Performance, and Memory in Flies. Curr Sleep Med Rep 1, 47-54 (2015).
Krishnan HC, Gandour CE, Ramos JL, Wrinkle MC, Sanchez-Pacheco JJ, Lyons LC. Acute sleep deprivation blocks short- and long-term operant memory in Aplysia. Sleep 39, 2161-2171 (2016).
Chandra R, et al. Sleep is required to consolidate odor memory and remodel olfactory synapses. Cell 186, 2911-2928 (2023).
Ardiel EL, Rankin CH. An elegant mind: learning and memory in Caenorhabditis elegans. Learn Mem 17, 191-201 (2010).
Mello CC, Kramer JM, Stinchcomb D, Ambros V. Efficient gene transfer in C. elegans: Extrachromosomal maintenance and integration of transforming sequences. EMBO J 10, 3959-3970 (1991).
Brenner S. The genetics of Caenorhabditis elegans. Genetics 77, 71-94 (1974).

There is NO Competing Interest.

Wangetal.Supplementaltable1relatedtoFig.1.xlsx
Supplemental Table1
Wangetal.Supplementarytable2.docx
Supplemental Table2
Wangetal.Supplementarytable3.docx
Supplemental Table3
SupplementaryFigures.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Sensory Plasticity Caused by Up-down Regulation Encodes the Information of Short-term Learning and Memory

Status:

Version 1

Abstract

Figures

Introduction

Results

Neuropeptidergic FLP-14/FRPR-10 signaling is required for salt chemotaxis learning

ASER-RIS circuit engages in SCL

FLP-14/FRPR-10 signal pathway in RIS mediates ASER sensatory plasticity in SCL process

Neuropeptide PDF-2 released from RIS up-regulates SCL and ASER sensory plasticity

ASER sensory plasticity encodes the information of the short-term salt chemotaxis learning

Discussion

Materials and Methods

Molecular biology

Behavioral assays

Genetic and chemogenetic manipulations of tested neurons

Calcium imaging

Confocal fluorescence imaging

Locomotion analysis

Neuropeptide-receptor interaction prediction with Alphafold2

Data analyses

Declarations

Declaration of interests

Data availability

References

Additional Declarations

Supplementary Files

Status:

Version 1