Harnagel, Audrey et al. (2021) International Worm Meeting "Identifying Triggers for Pathogen Learning in C. elegans"

Harnagel, Audrey, Bargmann, Cori

[International Worm Meeting, 2021]

The ability to discriminate between nutritious and harmful food is essential to survival. As a result, learned avoidance to harmful food sources is conserved from invertebrates to humans. The mechanisms enabling the nervous system to associate sensory cues from a food source with an internal state of sickness to trigger aversive memory formation remain elusive. After prolonged exposure to pathogenic food, C. elegans can learn to avoid the pathogen upon subsequent encounter1. This learned aversion requires infection; non virulent forms of bacteria are not sufficient for memory formation. In response to exposure to pathogenic food, serotonin is induced in a pair of sensory neurons called ADF and remodels downstream circuits2. Learned aversion to pathogen requires serotonin signaling from ADF, suggesting that ADF serves as a site of integration for detecting bacterial cues and internal sickness caused by the pathogen. We seek to understand how internal state changes the coupling between sensory activation and serotonin release in ADF neurons. As a first step, we are using calcium imaging to examine ADF responses to bacterial cues in both naive and pathogen-exposed animals. ADF responds robustly to conditioned media from both pathogenic and non-pathogenic bacteria in a dose-dependent fashion, and ADF activity can be modulated by previous odor history. We are screening mutants using these quantitative parameters to assess ADF responses to direct chemosensory stimuli, indirect signaling from other sensory neurons, and signaling from non-neuronal tissues indicating bacterial infection. Our goal is to uncover cell-biological mechanisms through which ADF neurons mediate learned pathogenic behavior in C. elegans. 1. Zhang, Y., Lu, H., & Bargmann, C.I. (2005). Pathogenic bacteria induce aversive olfactory learning in Caenorhabditis elegans. Nature, 438(7065), 179. 2. Morud, J., Hardege, I., Liu, H., Wu, T., Basu, S., Zhang, Y., & Schafer, W. (2020). Deoprhanisation of novel biogenic amine-gated ion channels identifies a new serotonin receptor for learning. bioRxiv. doi: https://doi.org/10.1101/2020.09.17.301382.

Levy, Sagi et al. (2021) International Worm Meeting "An adaptive-threshold mechanism for odor sensation and animal navigation"

Bargmann, Cori, Levy, Sagi

[International Worm Meeting, 2021]

Identifying the environmental information and computations that drive sensory detection is key for understanding animal behavior. A variety of different coding strategies have been proposed to mediate sensory detection across organisms and systems, including readouts of the signal absolute level, derivatives, or fold change, low-pass-filtered signal variables, and Linear-Nonlinear models. Differentiating between various models can be challenging without a deliberate systematic experimental design. Here we measure responses in the C. elegans olfactory neuron AWCON over a wide range of stimulus conditions. We find that previous sensation models could only match subsets of experimental observations. We formulate an alternative adaptive concentration threshold model in which sensory activity is regulated by an absolute signal threshold that continuously adapts to odor history. The activation threshold is extracted from a non-linear function of the signal followed by a low pass filter. The model fits the measured sensory threshold and latency over a broad stimulus range and accurately predicts sensory activity and probabilistic behavior during animal navigation in odor gradients. At a molecular level, the rate of the threshold adaptation is regulated by EGL-4, a cGMP-dependent protein kinase. The adaptive-threshold model generality was demonstrated by predicting activity of larval zebrafish optic tectum neurons in response to looming visual stimuli. Theoretical analysis shows that the adaptive concentration threshold model is better than the derivative and fold change models in filtering stimulus noise, allowing reliable sensation in fluctuating environments. Our model unifies previous sensation models under one mechanism and demonstrates an efficient encoding sensory mechanism that reconcile apparent tradeoffs between responsiveness, noise filtering, accurate detection and fast response speed.

Jin, Xin et al. (2013) International Worm Meeting "Studying the neural circuits of food choice imprinting in C. elegans."

Pokala, Navin, Jin, Xin, Bargmann, Cori

[International Worm Meeting, 2013]

Imprinting, a special form of learning during an early developmental stage, can generate a long-lasting memory and change animal behavior. We discovered a novel behavior in which early exposure of larval C. elegans to pathogen elicits a long-term aversion that is maintained for days, which we call food choice imprinting. It has been previously reported that adult C. elegans can learn to avoid pathogen Pseudomonas aeruginosa (PA14) after a single 6-hour exposure, but will lose the memory after 12 hours and return to the same preference as naive worms (Zhang et al., 2005). We modified this learning assay by performing pathogen training of newly hatched C. elegans larvae. Remarkably, larval-trained (henceforth, imprinted) animals retain this food choice memory days later, showing aversion to pathogen even as aged adults. To dissect the neural circuits of food choice imprinting, we used an inhibitory chloride channel to disrupt neuronal activity at different developmental stages, asking whether candidate neurons are required for memory formation during training, memory consolidation, or memory retrieval. In initial experiments, we have found that inhibiting the AIB interneuron during larval training disrupts performance, whereas inhibiting AIB during adult training does not affect adult learning. This result suggests that AIB is required to form a memory of the imprinted food choice, but not required for avoidance learning in adults. Through a combination of genetic and functional methods, we hope to elucidate the molecular basis of food choice imprinting in C. elegans. Reference: Zhang, Y.; Lu, H.; Bargmann, C. I., Pathogenic bacteria induce aversive olfactory learning in Caenorhabditis elegans. Nature 2005, 438 (7065), 179-84.

Albrecht, Dirk R et al. (2009) International Worm Meeting "C. elegans behavior in dynamic microfluidic environments."

Bargmann, Cori, Albrecht, Dirk R

[International Worm Meeting, 2009]

Neuronal circuits that govern goal-directed behaviors such as chemotaxis are highly interconnected, suggesting that their emergent functions may not be evident from the properties of individual neurons or connections. To study the underlying neural computation and dynamics, an engineering approach would systematically quantify output responses to many precise inputs under many circuit perturbations. While C. elegans locomotory behaviors are easily measured, and genetic tools enable precise circuit modification, the presentation of input stimuli in typical agar-plate assays is often poorly controlled. To address this limitation, we developed microfluidic liquid-filled arenas that enable the study of freely-moving animals in highly precise and dynamic microenvironments. We first optimized arena geometry to mimic C. elegans crawling motion, speed, and behavioral responses on agar surfaces. Microfluidic features create controlled liquid gradients that span several cm for population behavior or change sharply across the animal (<50 micron), and remain stable for hours or change rapidly within seconds. The transparent arenas are compatible with light-based neural control (via genetically-encoded rhodopsins) and fluorescent readouts of neural activity. We are characterizing wildtype C. elegans responses to complex spatial and temporal odorant patterns (steps and ramps) to understand how modulation of specific behaviors (e.g., speed, types of turns) influences chemotaxis strategy. Similar studies of genetic mutants with disrupted neurons or neuronal connections are revealing the role of perturbed information flow in directing these behaviors. For example, we found new behaviors (gradient-directed turning), new circuit pathways (glutamate-independent speed regulation), and strong phenotypes in subtle neuromodulatory mutants. Overall, the vast improvement in stimulus control in these microfluidic arenas enables new studies to understand the flow of information in neural circuits governing behavior.

Larsch, Johannes et al. (2013) International Worm Meeting "Maintaining sensitivity: dissecting sensory adaptation using high-throughput in vivo calcium imaging."

Bargmann, Cori I., Larsch, Johannes, Albrecht, Dirk R.

[International Worm Meeting, 2013]

Sensory neurons face a dilemma: Biologically relevant changes in stimulus strength are often small compared to the possible range of stimulus intensities. One solution to this problem is sensory adaptation, in which the sensory neuron continuously adjusts its dynamic range based on stimulus history to maximize sensitivity while avoiding saturation. During chemotaxis, C. elegans relies on known chemosensory neurons to detect odors over at least five orders of magnitude in concentration, suggesting a sophisticated molecular machinery to keep ongoing sensory activity within dynamic range. To probe adaptation in C. elegans, we wanted to deliver a broad range of odor stimulus concentrations and patterns while monitoring neural responses. To this end, we developed a high throughput system for in vivo calcium imaging that allows us to record up to 20 animals simultaneously for hours, enabling quantitative mapping of sensory receptive fields and their adjustment to constant or changing stimulus levels. We observed odor-evoked calcium dynamics in AWA sensory neurons across a million-fold range of concentrations. Neurons adapted upon repeated stimulation with odor pulses on two different time scales: fast inactivation of the calcium transient during each odor pulse and slow adjustment of response magnitude and dynamics across repeated pulses. We have screened mutants and pharmacological interventions for effects on odor-evoked calcium transients and their adaptation, seeking to define mechanisms for odor sensing and fast vs. slow adaptation. Paradoxically, several chemotaxis mutants with structurally defective cilia (che-2, che-3, osm-6) had stronger and more slowly adapting calcium responses to odor than wild-type animals suggesting that IFT cilia genes are more important for rapid adjustment of sensory sensitivity than for primary transduction in AWA.

Shinkai, Yoichi et al. (2009) International Worm Meeting "Molecular analysis of the integration of two sensory signals in C. elegans."

Tsunozaki, Makoto, Ishihara, Takeshi, Bargmann, Cori, Shinkai, Yoichi

[International Worm Meeting, 2009]

Environmental information is received by sensory neurons as sensory signals and subsequently those signals are integrated in the central nervous system. The sensory integration is a process of an adaptive response to the environment. If the sensory information is improperly integrated, the resulting abnormal outputs caused behavioral defects and learning impairments. Despite its importance, little is known about the mechanisms at the molecular level. To provide insights into the molecular mechanisms of the sensory integration, we performed a genetic screen to identify genes required for the sensory integration. In the mutant screening, an attractant diacetyl and a repellent copper ion were used. When a copper ion barrier exists between diacetyl and worms, worms must cross the copper ion barrier to reach the diacetyl spot. Mutants deficient in gcy-28, which encodes a receptor-type guanylate cyclase, less frequently crossed the barrier than wild type animals. In addition, gcy-28 mutants show the defect in the salt chemotaxis learning, which is the form of behavioral plasticity induced by paired stimuli, starvation and NaCl. To investigate the genetic interaction between gcy-28 and other genes involved in the sensory processing, we employed double mutant analyses in the sensory integration and in the salt chemotaxis learning. HEN-1/SCD-2 pathway regulates the sensory integration and the salt chemotaxis learning, whereas the insulin pathway regulates the salt chemotaxis learning. gcy-28;scd-2 double mutant and gcy-28;ins-1 double mutant showed stronger defects than each single mutant in both the salt chemotaxis learning and the sensory integration and in the salt chemotaxis learning, respectively. These results suggest that multiple pathways are involved even in a simple kind of sensory information processing. gcy-28 encodes four isoforms (gcy-28.a-d) (Tsunozaki et al., 2008). GFP reporter analyses suggested that gcy-28.d isoform was expressed mainly in AIA interneurons. The expression of GCY-28.d by a promoter that drives the expression only in AIA interneurons was sufficient to restore the phenotype of the gcy-28.d mutant tm3028 and partially rescued the defect of the mutant devoid of all gcy-28 isoforms. Our results suggest that gcy-28 regulates the sensory integration in multiple neurons including AIA interneurons, and provide insights into how sensory information processing is regulated in the nervous system.

Daghar, Hiba et al. (2021) International Worm Meeting "Developing C. elegans models for SRD5A3-CDG and Cori rare congenital diseases"

Daghar, Hiba, Parker, J Alex, Doyle, James J

[International Worm Meeting, 2021]

Rare congenital diseases are in urgent need of therapies. A simple animal such as the nematode Caenorhabditis elegans (C. elegans) may provide insights into the investigation of these diseases. Because of its defined genome and because many genes are highly conserved between C. elegans and humans, it is possible to investigate the underlying pathways of these diseases. In this study, we characterize two rare congenital diseases: SRD5A3-CDG, a congenital disorder of glycosylation type 1q and Cori disease, a glycogen storage disease type III. Current literature mainly shows clinical data concerning these diseases. SRD5A3-CDG disease is caused by a mutation in steroid 5 alpha-reductase 3 gene (SRD5A3), leading to significant psychomotor, cognitive and visual impairments. C. elegans has an orthologue of SRD5A3 called B0024.13 along with deletion (B0024.13(deletion)) and point mutation (B0024.13(W6X)) resulting in viable animals suitable for experimentation. B0024.13 C. elegans strains show significant motor impairments, neurodegeneration as well as neurodevelopmental delay. Cori disease is caused by a mutation in AGL gene coding for glycogen debranching enzyme which is involved in the glycogenolysis pathway. Clinically, Cori patients have significant hyperplasia of the liver, hypoglycemia, mental retardation and myopathy. The C. elegans orthologue of the Human AGL gene is agl-1. Our work with whole gene deletion (agl-1 (deletion)) and point mutation strains (agl-1 (W1044X), agl-1(S1444R)) showed significant movement impairments, neurodegeneration, and interestingly, a glycogen accumulation phenotype which make them suitable for drug screening experiments. An update of our work will be presented.

O'Halloran, Damien M. et al. (2011) International Worm Meeting "Contribution of cyclic nucleotide gated (CNG) channel subunits to behavioral plasticity in C. elegans."

Yu, Yawei, L'Etoile, Noelle D., Brueggemann, Chantal, Chen, Tsung-Yu, Zhang, Xiao-Dong, Altshuler-Keylin, Svetlana, O'Halloran, Damien M.

[International Worm Meeting, 2011]

The Caenorhabditis elegans AWC neurons are responsible for sensation of a range of attractive volatile odors (Bargmann et al., 1993). Prolonged odor exposure in the absence of food leads to reversible decreases (adaptation) in the animal's attraction to that odor (Colbert and Bargmann, 1995; L'Etoile et al., 2002; Nuttley et al., 2002; Kaye et al., 2009; O'Halloran et al., 2009). It has been shown previously that the odor specificity of adaptation is determined by the feeding status of the animal (Colbert and Bargmann, 1997). That is, if a well-fed worm is exposed to benzaldehyde for a sustained period, it will adapt to both benzaldehyde and isoamyl alcohol (both sensed by AWC), this process is termed cross adaptation. In contrast, an unfed (starved) worm will adapt to benzaldehyde and its response to isoamyl alcohol will remain intact. The TAX-4 and TAX-2 cyclic nucleotide-gated (CNG) channel subunits are required for AWC-mediated olfactory responses (Coburn and Bargmann, 1996; Komatsu et al., 1999). TAX-4 is an a subunit that can form homomeric channels while TAX-2 is a b subunit that requires TAX-4 to form a functional channel (Coburn and Bargmann, 1996; Komatsu et al., 1999). C. elegans encodes two additional predicted a subunits, CNG-1 and CNG-3 (Cho et al., 2004a; 2004b; Coburn thesis, 1996). Here we report that CNG-1 is required in AWC to promote short-term (30-mins) cross adaptation between benzaldehyde and isoamyl alcohol. The ability of food to induce this cross adaptation is also dependent on the ASI sensory neurons. We also demonstrate that CNG-3 is required in AWC for adaptation to short-term (30-mins) exposures of odor. By using FRET, Bio-molecular Fluorescent Complementation assays, and genetically encoded calcium imaging we find that TAX-2::TAX-4::CNG-3 channels may promote short-term adaptation in AWC. Our data suggests the TAX-2::TAX-4::CNG-3 channel adopts a 3a:1b stoichiometry. By examining the electrophysiology of the CNG subunits in AWC, we also demonstrate that fast closing dynamics appear critical for proper short-term adaptation responses. Taken together our findings provide more understanding into the mechanisms and circuitry of how CNG channels shape olfactory plasticity.

Jang, Heeun et al. (2011) International Worm Meeting "Internal state and sex regulate an ambivalent circuit for pheromone responses."

Bargmann, Cori, Jang, Heeun, Sengupta, Piali, Kim, Kyuhyung, Butcher, Rebecca

[International Worm Meeting, 2011]

Animals respond flexibly to environmental cues based on internal and external context. Cocktails of several ascarosides, the components of C.elegans pheromones, are repulsive to wild-type hermaphrodites from the solitary strain N2 but are less repulsive or even mildly attractive to social hermaphrodites with reduced activity of the npr-1 neuropeptide receptor gene. We have examined the circuit mechanisms underlying these alternative behavioral responses to pheromones using behavioral assays and in vivo imaging with genetically encoded Ca2+ sensors. We found that the C9 ascaroside (also called ascr#3) evokes strong avoidance behaviors in N2 hermaphrodites. This avoidance is mediated by the ADL chemosensory neurons, which directly sense C9. Interestingly, pheromone attraction in npr-1 mutant hermaphrodites also requires ADL, which cooperates with the previously described ASK sensory neurons to detect blends of ascarosides C9 and C3 (also called ascr#5) that are not individually attractive. ADL and ASK both belong to a hub-and-spoke circuit of neurons connected through gap junctions to the central hub neuron RMG, which is regulated by npr-1. Attraction to C9 and C3 requires RMG synaptic activity, but avoidance of C9 does not. Conversely, avoidance of C9 requires ADL synaptic activity, but attraction to C9 and C3 does not. Our results suggest that the balance between avoidance and attraction can be regulated by changing the relative strengths of ADL chemical synapses compared to the ADL/ASK-RMG gap junction circuit. A parallel mechanism generates sexual dimorphism in pheromone responses. Males exhibit reduced avoidance of high C9 concentrations, and their ADL sensory neurons are less responsive to C9 than those of hermaphrodites. The effects of sexual dimorphism are additive with the effects of npr-1, suggesting that these pathways function independently. Modulation of neuronal responses by neuropeptide signaling and sex may permit small neuronal circuits to maximize their adaptive responses and behavioral outputs.

N A Dunn et al. (2002) West Coast Worm Meeting "Simulating neural networks for spatial orientation in C. elegans"

S R Lockery, J S Conery, N A Dunn

[West Coast Worm Meeting, 2002]

In C. elegans, spatial orientation behavior in a chemical gradient (chemotaxis) involves a series of turns (pirouettes) whose frequency is modulated by the change of attractant concentration. Ablation of identified neurons has delineated a candidate neural network for chemotaxis in C. elegans (C. Bargmann, UCSF, unpublished data). The aim of our research is to generate testable models of how this network computes behavioral state and consequently turning frequency in response to changes in concentration.